Jan Vansina

Art History
in

Africa

USED
RAMS BOOK STORE
NO EXCHANGE Of) REFUNDS
W/0 HZt TAG f> RECEIPT

>^

\0D

J.

VANSINA

ART HISTORY IN AFRICA

AN INTRODUCTION TO METHOD

Drawings by C. Vansina

BOOK RETURN POLICY

1-FULL REFUND

FOR FIRST 10 DAYS OF CLh:

2-CASH REGISTER RECEIPT REQWRED

3-NO MARKS IN NEW BOOKS

RAMS BOOK STORE

Longman -S!
London and New York

Longman Group UK Limited

Longman House,

Bvirnt Mill, Harlow, Essex

and Associated Companies throughout

CM20

2JE, England

the world.

Published in the United States of America

by Longman

New

Inc.,

York

Longman Group Limited 1984

All rights reserved; no part of this publication

may be

reproduced, stored in a retrieval system, or transmitted

in any form or by any means, electronic, mechanical,

photocopying, recording, or otherwise, without either

the prior written permission of the Publishers or a
licence permitting restricted copying in the

Kingdom

United

Agency
7DP.

issued by the Copyright Licensing

Ltd, 33-34 Alfred Place, London,

WCIE

First published 1984

Fourth impression 1990

British Library Cataloguing in Publication

Data

Vansina, Jan

Art history in Africa.

L
L

Arts, African

Title

709'.6

N7380

ISBN D-sfiE-b^3ba-b
Library of Congress Cataloging in Publication Data
Vansina, Jan.

Art history in Africa.

(Longman

studies in African history)

BibUography:

p.

Includes index.

L Art, African L
N7300.V36 1983

ISBN
ISBN

Title.

IL Series.

709'.6

82-21644

0-582-64367-8
0-582-64368-6 (pbk.)

Produced by Longman Group (FE) Ltd

Printed in

Hong Kong

CONTENTS
Preface

viii

List of plates
List of figures

xii

Acknowledgements

CHAPTER

1:

xiv

INTRODUCTION

African art, arts of Africa

Art and its history
African geography
The dawn of art in Africa

1
1

4
6
7

The oikownene
The regional traditions
West Africa

10

Central Africa

14

East Africa and southern Africa

The study of African art

17

CHAPTER

21

2:

11

IDENTIFICATION

Labels, objects and documents

itself: description

The work

Forgeries and their ilk

Author and place of origin

19

21

24
26
27

Physical dating

33
36

Catalogues

40

Dating

CHAPTER

3:

SOCIETY,

THE MOTHER OF ART

41

Use

41
42

Patronage

44

Specific social context

47
50
52

Object and society

Artists

and workshops

Function

Documentation of

social context

53

CHAPTER

MEDIA AND TECHNIQUES

4:

Architecture
Additive sculpture: metalworking
Additive sculpture: clays and related media
Subtractive sculpture: wood
Subtractive sculpture: ivory, stone

Mixed media
Painting and drawing
Textiles and other fabrics

Ornamenting the body: scarification, painting

Media, techniques and the history of art

56
57
61

64
"5
67
"
"8
70
73
73

STYLE

78

Conventions
Morphological analysis
Atelier and attribution

78

CHAPTER

5:

Stylistic seriation

Shape

QO
"^

time

CHAPTER

o^
^^
"2

6:

THE INTERPRETATION OF ICONS

101

Icon and concept

motif

1^1

Theme and

1^^

Interpretation

1^"
1 in
^^"
112
H'^
117

Art language
The complexity of meaning
Decoration: art without a statement
Theme and motif in history

CHAPTER
The

7:

visual arts

CULTURE AND ART

compared

Visual and performing arts

Repertoire and culture
Aesthetics

Dynamics of

CHAPTER
The

art

8:

and culture

THE CREATIVE PROCESS

artist as creator

121
122

126
129
130
133

136
136

Originality

1*^1

Replications

l'^2

145

Drift

1"^

Prime works
Creativity, social

and cultural change

154

CHAPTER 9: THE CREATIVE PROCESS: FOREIGN INPUTS

Diffusion: the

Demand

means
works of

for foreign

art

Distributions

Reconstructing diffusion
Assessing the evidence

159
159
161
164
168
170

WIDER PERSPECTIVES

174

Formal frameworks
Framework by an area
Tree models: Nigeria from Nok to Yoruba
Western Nigerian style relationships
frameworks

174
174
177
178
185
193

IN HISTORY

196

CHAPTER

An

10:

operational model: streams of tradition

The

failure of formal

CHAPTER

11:

ART

Stone mansions at Lamu in the eighteenth century

Art objects as sources
Art in economic and social history
Art and intellectual history
The relevance of art to history

196
201
203
205
211

References and further reading

Index

214
225

PREFACE
a child artworks enthralled me. They seemed to struggle to speak, to
whisper about their times, their landscapes, the people that roamed the roads
in those days when everything was as different as the characters in a fairy tale
are from real people. But the eye could never understand all by just gazing at
this window into a world as far away as the mirror image at the bottom of a deep
well. This feeling never left me completely, but I came to understand that
works of art must be helped, not by the imagination, but by the retrieval of as
full an historical setting as research can uncover.
This is as true for Africa as for any other continent. And yet how often are
works of African art viewed as if they came from nowhere, as if they did not
raise questions about where and when they were created? Even today most
commentaries deal only with ethnographic context or parallel, and with form.
How often they therefore remain flawed and shallow! Still, on the whole the
concerns, goals and tools of art history are beginning to appear in the study of
art in Africa, albeit as timidly as the first crocuses appear after the tide of
winter. And yet an awareness of the historical setting needs to permeate all
research about art, if art is to make sense at all. The greatest masterpiece is
timeless only because it captures the evanescent spirit of its own lime. Art
cannot properly be understood at all without history. This book has grown out

As

of that conviction.
Several scholars have read and generously commented on this manuscript
in earlier forms. It is a pleasure to thank Professors J. Watrous and F. Neyt, Dr
D. Henige who read and commented on the whole manuscript as well as
Professor S. Feierman who shared his thoughts about the chapters dealing
with society and culture. I also owe a debt of gratitude towards the sometimes
intemperate reviewer for Longman whose frankness was enhanced by his
anonymity. The book, I hope, is better because of his comments. So thanks to

have been greatly assisted in the background work for this book by
Dr B. Fulks, who helped me with the teaching of a
course on art history in Africa. The manuscript became clear thanks to the
superior skills of Mrs Rosso who typed it. But the text is only part of such a
book. The individuals and the institutions who graciously allowed me to
reproduce their photographs in this book should not be forgotten in any
acknowledgements. The list of illustrations records their names. The Vilas
Research Fund of Madison (Wisconsin) is also gratefully remembered. Its
support gave me time to think and write while providing some of the necessary
travel funds as well.
And then I come to the collaborators. First Claudine, whose talents
produced all the drawings, diagrams and maps. She has been the most
J. P. also. I

the assistants, especially

viii

PREFACE
persistent
are

commentator and one endowed with the

gift

of seeing.

Then

there

my parents, both painters, who taught me art history. My father was also an

art historian

whose formidable knowledge was matched only by

book is dedicated to his memory.

his genius for

interpretation. This

J. Vansina

IX

OF PLATES

LIST

2.1

2.2

2.3

Divination board, Ardra, Republic of Benin. Ulmer Museum

Figure, Kina. Museo preistorico ed etnografico, Rome
Mask, Bukoba, Tanzania. Etnografisch Museum, Antwerp
Divination board, Yoruba, Nigeria. Museum fiir Volker-

30

kunde, Berlin-Dahlem

Museo

preistorico ed

2.5

Ivory bowl, Sherbro area. Sierra Leone.

etnografico, Rome
Tellem figure. L. Wunderman

2.6

Tellem

3.1

Amsterdam
Crown or shrine?, Loango, Angola. Musee de I'homme,

2.4

Koninklijk

figure.

Instituut

35
37

voor

de

Tropen,

Nomoli

3.3

Kuba

figure. Sierra Leone. British

royal

Midden

Museum

drums. A. Scohy. Koninklijk

Museum

voor

48

Afrika, Tervuren

3.4

Lele village drum. Koninklijk

3.5

Emblem

Museum

voor Midden Afrika,

49

Tervuren

4.4

with a hand, near Sankuru. Museum fiir

Volkerkunde, Berlin-Dahlem
King Munza's reception hall. G. Schweinfurth
Minaret, Agadez, Niger. S. Denyer
Dolls, Cape Town. Museum fiir Volkerkunde, Munich
Bronze vessel, Igbo Ukwu, Nigeria. National Museum, Lagos

4.5

Carver, Bolony.

4.6

Seated figure,

4.7

Tomb

4.1

4.2
4.3

Midden

J.

Zaire. Koninklijk

Museum

66
69

Webb

4.8
5.1

Horned mask with

5.2

University of Pennsylvania, Philadelphia

Reliquary figure, Taki. Museo preistorico ed etnografico,

bird. Central Ivory Coast (Guro).

76

Museum,

Rome
5.3

5.4
5.5

Ancestor figure, Gabon (Fang). Museum fiir Volkerkunde,

Berlin-Dahlem
Ancestor figure, (Shaba) Hemba, Zaire
Kabila figure, Shaba. Koninklijk Museum voor Midden
Afrika, Tervuren

Acknowledgements: the first mention refers to the photographer, the second to the owner
from the photographer or if not a pubUc building.

different

58

60
62
63

voor

Afrika, Tervuren

Amen

54

65

Vansina

Kambundi,

Khopshef, Thebes. T.
Fagade, Abba Lebanos. P. Curtin
of

38

42
43

Paris
3.2

23
28

79
80
86
88

90
it

LIST
Zaire. Etnografisch

OF PLATES

Museum, Amwerp

5.6

Charm, East Kasai,

5.7

Benin 'Early Style' head. British Museum

Chibinda Ilunga figure. Museum fiir Volkerkunde, Berlin-

6.1

6.3

Figure of a hunter, Lower Niger. British Museum

Family scene, Karkur Talh, Libya. J. Van Noten
Koninklijk Museum voor Midden Afrika, Tervuren

6.4

Family scene, Sefar, Algeria. J-D. Lajoux

6.5

Tomb

6.6

Female

6.7
6.8

of

Menna, Shaikh 'Abd

Keenge.

figure,

J.

al

105

and

Qurnah, Thebes

Vansina

Madonna Hodegithria, Ethiopia. Museum

Munich
Horus and Sebk. Musee du Louvre, Paris

fiir

109
112

H^
118

7.3

7.4

Main

8.1

Minaret, Quttubiya, Marrakush, Morocco.

8.2

Commemorative

7.2

107
107

Volkerkunde,

Minaret, Hasan mosque, Rabat, Morocco. P. Curtin

Maskers cartwheeling, Tsayi, Congo. M-C. Dupre
Tusk, Loango, Congo/ Angola. P. Curtin. Walters Art Gallery,

7.1

97
104

Dahlem
6.2

91

125

128
1^1

Baltimore
palace and enclosure,

Koninklijk

Zimbabwe.

P. Garlake
P. Curtin

(Ndengese), near Dekese, Zaire.

voor Midden Afrika, Tervuren
Kuba. Museum fur Volkerkunde,

134
143

figure

Museum

8.3

Cephalomorph

8.4

Berlin-Dahlem
Statue of king Shyaam aMbul aNgoong. British

cup,

149
150

Museum

152
153

vessel. Yale University Art Gallery

8.5

Kuba

8.6

Portal, Sultan Hasan's complex, Cairo, Egypt. E.

155

9.1

Raffia pile cloth, former

Part figure, Jemaa, Nigeria. National

166
181
182
188

10.1

Bohm
kingdom Kongo. Ulmer Museum

Nigeria. British

Museum, Lagos

Museum

10.2

Crowned head,

10.3

Ceramic bust. Inner Niger delta, MaU. B. de Grunne

Maskers, Mossi, Upper Volta. L. Frobenius. Frobenius

10.4

Institut,

Ife,

1^0

Frankfurt

11.4

Head, Lydenburg, Transvaal. Natal Museum

Inner wall, Lamu mansion, Kenya. J. de Vere Allen
Goldweights, southern Ghana. T. Garrard
Female figure, Mapey, Zaire. Koninklijk Museum voor
Midden Afrika, Tervuren
Charm figure, Kuba, Zaire. Museum, University of

11.5

Pennsylvania, Philadelphia
Statue of a queen of the Holo.

10.5
11.1

11.2

11.3

11.6

Doorpanel, Lemba,
Berlin-Dahlem

Zaire.

J.

fiir

198

202

206
207
209

Hautelet

Museum

192

Volkerkunde,
212

XI

LIST

1.1

OF FIGURES

LIST

OF FIGURES

10.1

10.2

Major towns and sites in western Nigeria

Head, Owo
Style and Nigerian art history
Four-legged pots along the Upper Niger before A.D. 1500
House, Lamu (plan)

10.3

10.4
10.5
11.1

tree

model: western Nigerian styles

179
180
183
186
189
197

xni

ACKNOWLEDGEMENTS
The

publishers are grateful to the following for permission to reproduce

text: J. de V. Allen for pi. 11.1; Erwin Bohm for pi. 8.6;
Collection Troppenmuseum, Amsterdam for pi. 2.6; Comte B. de Grunne for

photographs in the
pi.

10.3;

Courtesy of the Trustees of the British

Museum

for pis 3.2

(Museum of Mankind), 5.7 (Museimi of Mankind), 6.2 (Museum of Mankind),

8.4 (Museum of Mankind), 10.1 and 10.2 (Museum of Mankind); MarieClaude Dupre, Centre National de la Recherche Scientifique, Ouagadougou
for pi. 7.2; Etnografisch Museum, Antwerp for pis 2.2 and 5.6; Werner
Forman Archive for pis 4.2 and 4.4; Frobenius Institute for pi. 10.4; Peter
Garlake for pi. 7.4; T. Garrard for pi. 11.2; The Johns Hopkins University
and 8.8; Koninklijk Museum voor midden Africa, Tervuren,
Belgium for pis 3.3, 3.4, 4.6, 5.5, 6.3, 8.2 and 11.3; M. Jean-Dominique
Lajoux for pi. 6.4; Musee de I'homme for pi. 3.1; Musee de Louvres for pi.
6.8; Museo Nationale Preistorico ed Etnografico for pis 2.1, 2.4 and 5.2;
Museum fur Volkerkunde W. Berlin for pis 2.3, 3.5, 5.3, 6.1, 8.3 and 11.6;
Natal Museum for pi. 10.5 (Tim Maggs); F. Neyt for pi. 5.4; Societe d'Arts
Primitifs for pi. 11.5; Staatliches Museum fur Volkerkunde, Miinchen for pis
4.3 and 6.7; Ulmer Museum, Ulm, W. Germany for pis 1.1 and 9.1
(Collection Weickmann); University Museum, Philadelphia for pis 5.1 and
11.4; University of Wisconsin Press/Dr. Thompson Webb for pi. 4.7;
J. Vansina for pis 4.5, 6.6; Walters Art Gallery for pi. 7.3; L. Wunderman
for pi. 2.5; Yale University Art Gallery for 8.5; pi. 4.1 was taken from the
publication 'The Heart of Africa' by G. Schweinfurth published by Sampson
Low, 1878; pi. 6.5 appeared in Wreszinski, 'Atlas Zur' published by J.
for pis 4.8, 7.1

Hinrichsen Leipzig, 1923 Atlag hcher Kulturgeschichte.

The

cover photograph was kindly supplied courtesy of the Trustees of the

Museum (The Museum of Mankind).

British

XIV

CHAPTER ONE

INTRODUCTION
AFRICAN ART, ARTS OF AFRICA
art' is the label usually given to the visual and plastic arts of the
peoples south of the Sahara, especially those of western and central Africa. Not
only have students of African art limited themselves to a portion of the
continent but they have been primarily concerned with the aesthetic appeal of
sculpture and a description of the uses and functions of the objects in an
ethnographic context. Thus the historical evolvement of the art forms, even
the sculptural forms, has not been a subject of sustained research and,
furthermore, other kinds of an have received scant attention. Thus defined,
'African art' is not the Art of Africa. Northern Africa, almost half of the

'African

continent in size, has been excluded from those studies because its arts clearly
belong to widely flung traditions centered on the Mediterranean and the
worlds of Christianity and Islam. These traditions can be called oikoumenical,
from a Greek expression (oikoumenikos) meaning 'the whole world'. They
transcend local and regional cultures over many lands. The contrast with the

We

cannot amputate half of Africa

regional traditions of art elsewhere is great.
art'. Moreover, by its
'African
remains
what
of
portion
call
a
then
and
emphasis on sculpture even among regional traditions, the artistic expressions
of eastern and southern Africa where sculpture is not the major form of artistic
expression in recent times are also slighted.
Because of this, the present book deals with art in Africa and its history. It
monographs to write
is not an art history of Africa. There are as yet not enough
art' have been
'African
of
field
such an art history, as too many scholars in the
of the
problems
historical
with
here
deal
we
But
pursuits.
allergic to historical

an expose of the approach to art

an introduction to the questions art
historians should ask about the objects of study and to the ways they should
follow when seeking answers to such questions. It applies the general
epistemology and methods used in the discipline to the specific situation of art
in Africa. It should also be of help in evaluating historical hypotheses made
about art in Africa. If this book becomes a stimulus to historical study, its
purpose will be fulfilled along with the hope that one day it will indeed be

art of Africa before

AD.

history in general as

it

1900. This

book

relates to Africa. It

is

is

possible to write an art history of Africa.

ART AND

ITS

HISTORY

term of western culture but a very inexact one. The threshold between
what may be judged a work of visual art and another kind of man-made object

Art

is

ART HISTORY
is

IN AFRICA

often a matter of dispute. For example, in our time, the distinction between

music and what noise has become frayed. A

we can say is that art
deals with form and expresses images or metaphors (Layton, 1981:4-15).
Yet an 'aesthetic' drive is universal. Everywhere and at all times people
have made objects or manufactured decorative patterns that are unnecessary
from the point of view of use. Even cooking pots are not entirely determined by
use. Their shapes vary from place to place over time and archaeologists use
them as prime determinators of 'culture'. There exists everywhere a need for
the formal expression of values by metaphorical means, an appeal to the senses
of sight and touch. The visual and plastic arts are means through which this
need can be satisfied. Any made-made object studied from the point of view of
form may be an art object, and form is a major concern of any study of the arts,
whether or not the objects will be lasting, whether or not the object was made
just to express form, whether or not the object is a man-made thing or merely
an embellishment of some other object, such as painting on the human skin.
Art historians also investigate iconography, the characteristics and meanings
of pictorial renderings or of symbols whose arrangements, and even specific
location in compositions, affect the form of the art or, conversely, the form
affects the iconography. They further study media and technologies as
methods and materials that allow a concordance between the artistic
conception and the aesthetic form of the work of art.
In practice our illustrations are mostly taken from architecture, especially
public architecture, sculpture and tainting or drawing, whether figurative,
stylized or decorative only. The illustrations like the text are but an
introduction to the field. Hence they represent such works of art as can be most
easily linked to historical concerns, leaving such art historical problems as

what was once considered

lapidary definition of art

Vlatel.l Divination hoard. Wood.

Width 34cm. Before 1659

to be

is

therefore meaningless. All

Usediniii divination. Ardra, Republic of Benin. U Inter .Museum.

Height 55cm,

INTRODUCTION
in East Africa or sculptured hairstyles in southern Angola aside
altogether, because those manifestations of art are ephemeral, and thus very
difficult to document in the past. If we consider a work of art as in Plate 1 1 , we

body painting

can

illustrate the types of questions

it

raises. Is this a tray?

from? When was it made? How was it made?

Where does

it

come

what it pretends to be? Is it

unique in its general form or one of several or only a copy of something else?
What was it used for? What did it mean? How does it fit in the whole of artistic
production in Africa and elsewhere? Who made it? For whom? A jumble of
questions pours out. The job of art history is to answer these and to do so in an
orderly manner, first identifying the object as to authenticity, place and time of
production, the artist, the manner of fabrication, the style, the meaning and
the socio-cultural context relating this to the whole culture. Then it examines
the idiosyncrasy of the object in comparison to others, that is, the conditions,
circumstances and quality of its creation both internal and external. Finally it
places the object in a general framework of the evolution of similar and related
art forms in Africa, and by inference, in the world. This book follows this order
Is

it

of asking questions throughout.

it is authentic. It came in the seventeenth
Ardra, a town on the coast of the present
from
Europe
century (before 1656) to
Republic of Benin (Dahomey). Its maker, the exact location of the workshop in
which it was made, and the date of manufacture are unknown. Our
information about it stems from a catalogue in 1659:

As

to the art object cited,

board carved with strange, marvelously rare and horribly

which the King in Ardra, who is a vassal of the great king
of Benin, and his most important officers and people of the same
province, use to employ for the sacrifices to their gods or fetishes and on
which they are wont to sacrifice to them. And this sacrificial board has
been desecrated* by the ruling king of Ardra himself and has been used by

sacrificial

devilish images,

him {Exoticophylacium:

52).

was a board used for divination. Ifa is the name for a god linked to it, for the
system of divination of which it was a tool, and for a cult whose priests are the
diviners. The face on the board may be Ifa's or Legba's, another god who tricks
people into offending the gods. Many details of its iconography are still not
understood. Most occur on other objects from the seventeenth to the
nineteenth and twentieth centuries. Ifa boards are still made in the general
area. It was probably not the first ifa board of its type to be made, although it is
the oldest of those recovered. 'Foreign' influence has not been detected in it. In
general it belongs to a great tradition of sculpture in western Nigeria labelled
'Yoruba art' of which it is the earliest dated specimen. Note that the object was
not found in a town inhabited by Yoruba speakers (Willett 1971).
Why should there be an art historical study of Africa for until recently
there have been doubts that it was possible or even desirable (Volavka 1979)?
First, perhaps, because art from remote times exists there. Some graphic
works survived from dates as remote as 25 000 B.C. in Namibia and 6000 B.C.

It

(Grimm, Duden). Presumably: 'desacralized'. Amended to

would mean 'commissioned' which would make more apparent sense in the context.

'Infesiiert' not in dictionaries

'Investiert'

it

ART HISTORY

IN AFRICA

and plastic arts have a long and complex past.

Secondly, works of art are always transformations of whatever kinds of realities
people experience. And, in a given time, the nature of the art of a culture
evolves as a response to realities - seen, felt or taught - which artists then
record, with as much perfection of form as their talents can summon, in objects
that may be descriptive or expressive, symbolic or decorative. As social and
cultural values of a people change with time, so do the substances, statements
in the Sahara. African visual

is history, and the more we can know about the

times and places, the better we are able to understand the
perceived realities of a people as expressed in their art.
The art history of African regional traditions, however, is still in its
infancy. There are various reasons, of which the one of most concern is that
only the merest fraction of art has survived, compared to the mass that was
once in existence. As this book will argue, however, we can still recover much
of the history of these traditions. The main reason for a lack of interest in the
history of these arts was simply the haphazard development of the field of

and

aesthetics of their art. This

art of successive

'African art' and the circumstances surrounding the 'discovery' in

Europe that

Africa had an art.

For a long period the study of regional arts in Africa were contained in the
expressions of a Western aesthetic response, as if it were a wonder of nature or
of outer space, an exotic specimen in the gallery of visual images in Europe.
Later, more serious students turned to the determination of style and the
position of the work of art in its social and cultural milieu, but - barring a few
exceptions - without considering any time scale, instead they used an
immobile 'ethnographic present' tense to describe the object in its context. It
was as if creativity in Africa had been frozen after some genesis when the
types of icons were crafted by the hands of some hero of a founding
myth. Clearly this approach will not do. It still deprives works of African art of
the full measure of attention and study they deserve if they are to be properly
understood. It is faulty because even the contexts described for some objects
may not apply to older works of art, created in circumstances very different
from the ones observed by anthropologists. An historical analysis is absolutely
necessary before a study of art in Africa, initiated by descriptions of style and

known

context, can be completed.

and a methodical approach to history are essential to the study of

any general history. It is essential because any
history cannot be called general if it does not include art, and because art is a
contemporary and authentic expression of the concerns of an age and a
community, while changes in art reflect changes in such concerns. Because
If history

art, art history is also relevant to

historians of Africa have also neglected the history of its art, the last chapter in
this book briefly discusses how art can contribute to the general enterprise of
historians.

For

art objects are often

primary evidence of times long past.

AFRICAN GEOGRAPHY
and the most massive in outline, forms
with nearby Asia and Europe the Old World. And an old world it is, as
mankind itself evolved in Africa and spread from it to the rest of the world. The
continent is deeply marked by its vegetation belts and its orography. From

Africa, the second largest continent

INTRODUCTION
north to south a belt of mediterranean vegetation fades into the largest desert of
the world, the Sahara, crossed only by the oasis ribbon of the Nile. Further
south grasslands, called the Sahel in West Africa, gradually covered with dry

#''0 tf

ART HISTORY

IN AFRICA

one moves south, give way to rain forest near the coast of West Africa
across central Africa. East Africa's elevations keep a wide corridor of
grassland of various sorts open to the south. At about lat. 4' south of the
Equator grasslands and dry forest appear again only to fade in more open
country further south and finally run into the deserts and extremely dry lands
forest as

and

all

of Namibia and the interior of Cape province. There, a mediterranean type of

vegetation suddenly appears again in a tiny portion of the continent around the
Cape proper.
Gradually the vegetation belts have shifted and the Sahara has not always
been a desert, but the rain forest has always been there. These facts are
important because the desert and the forest have acted not as absolute, but as
partial barriers to communication. Thus the shape of the continent and the
vegetation belts show that southern Africa was the most isolated from the rest
of the world, central Africa was cut off from the lands north by the forest and
the desert (after c. 2000 B.C.) separated West Africa from North Africa. East
Africa communicated with southern Africa, but the existence of a great
geological rift containing a string of major lakes separated it from Central
Africa and

hampered communications.

the continent provided their own highways and

was part of an Indian Ocean world in which the monsoons
regulated the seasons for trading. The currents and winds of the Atlantic
Ocean prevented any passage west of the Sahara until the caravel, that
space-shuttle of the Renaissance, was developed. When the door opened from

The oceans around

barriers. East Africa

c.

1450 onwards European ships could reach

and

all

of Africa's coasts, but not

lethal tropical diseases

its

unknown to

interior defended by deltas, waterfalls

Europeans. The Mediterranean in the north has served at all times as a link,
more than as a barrier. It was the heart of what was to become the oikoumene
around it. That the oikoumene encompassed all northern Africa and all
northeastern Africa as well as its eastern seaboard to Mozambique - the limit of
the monsoons - is therefore not surprising. Because the Sahara dried up only
gradually and last of all on its western flank, and because of its gold deposits,
West Africa never totally lost contact with the oikoumene, but still never

became

a part of

it.

has been that in certain areas

droughts were numerous and populations had to be nomadic or migrate often,
preventing the creation of bulky works of art. Thus sculpture and a complex
architecture were precluded in large portions of the interior of eastern Africa,
and in southwestern Africa.

One effect of the geography of the continent

THE DAWN OF ART

IN AFRICA

years and more before the first Pharaoh was enthroned,

hunters and pastoralists in the Sahara were both engraving and painting on
rock. Indeed, ancient Egyptian graphic art owes something to the great
Saharan tradition that both preceded it and ran parallel to it for most of its
history. Painting and engraving are typical for nomadic people, and were
probably practised all over the continent before the spread of settled life.
Ahhough rock art dated to very early times has only been found in southern

Some two thousand

Africa and the Sahara, later rock art

is

known from

almost anywhere where

INTRODUCTION
rocks were available. Apart from tiny sculptures in the Sahara and in southern
Africa, architecture and sculpture truly developed only with more sedentary
ways of life. Fishing, agriculture and some forms of husbandry allowed such
developments in places from about 7000 B.C., and then increased with the
spread of cereal agriculture and the domestication of animals a thousand years
later.
first developed for copper, bronze, gold and silver, existed by
Pharaonic
Age, c. 3000 B.C., and then spread from northern
the dawn of the
Africa to Mauritania where Akjoujt's mines were exploited between the ninth
and the fifth century B.C. By that date iron technology had become well
established. The smelting of iron may have first reached Africa with the
Phoenician colonies from c. 1000 B.C. onwards. By 700 B.C. at the latest their
technology had crossed the Sahara and had been refined at such foundries as
Taruga in Nigeria. By 500 B.C. iron was worked in Ethiopia and by the same
date people in the Great Lakes area of East Africa were melting iron at high

Metallurgy,

elaborated in India. Be that as

West and East Africans were
producing high carbon steel directly from the furnace. Not long after the turn
of the era, iron metallurgy had reached Natal (Van der Merwe 1980; Schmidt
1981). The sophistication and the independent innovations in African metal

temperatures, perhaps developing a process

it

may, from the

last

first

centuries B.C. onwards, both

technologies at such early dates should place historians of art on the alert. Even
if we have no works of art in metal dating directly from such remote times,
do have evidence for
there remains the probability that they were made.

We

complex ceramic sculpture called A/o/j in Nigeria from 700-500 B.C. onwards.
It was a mature art and one may think that sculpture was also practised in other
more perishable media as well.
By that time the separation between the oikoumene and the regional arts of
Africa began to appear. Nok flourished at the time of Herodotus (485^25
B.C.), whose reports show that the Sahara was then almost as dry as it is now,
although it still could be, and was, crossed by horse-drawn chariots. Most of
the Saharan populations had moved southwards before the onslaught of
drought, including perhaps some forebears of those who crafted the works of
the Nok tradition. The gulf between the Mediterranean and the continent
beyond had opened.

THE OIKOUMENE
At first the dominant arts of the: oikoumene were tied to political developments,
and their associated religions. The Pharaonic achievements first, then
Carthaginian and Greek colonies, later still the expansion of the Roman
Empire, all affected the spread and growth of the arts and provided subject
matters, uses and functions. Then from the first century A.D. onwards,
religion - less and less tied to empire - became the driving force.
Christianity reached African shores from its first generation of believers
onwards. After long struggles it was recognized by the edict of Milan in A.D.
313, the date conventionally assigned to the onset of Coptic art in Egypt.
Eastern Christian art spread to both Nubia and Ethiopia in the next centuries,
in both areas succeeding arts already influenced by theoikoumene for more than
half a millennium at least.

The flight of the prophet Muhammad from Mecca in

ART HISTORY IN AFRICA

INTRODUCTION

ART HISTORY

IN AFRICA

622 signalled the birth of Islam. The new religion and the new Empire, still
developing its own artistic expression, entered Egypt in 639. The first
conqueror of the Maghrib, 'Uqba ibn Nafi', founded the first mosque at
Qairawan in 670. But Muslim forces were not well entrenched in the Maghrib
before A.D. 710. From then on, however, Muslim art began to develop and
displayed major local expressions a century later. In Egypt, meanwhile,
masterworks of Islamic art had appeared even earlier. Islamic art spread with
the later expansion of religion and of Arabs over northern Africa, already
reaching the East African coast as well as portions of Madagascar from the
ninth century onwards, and West Africa in the eleventh century A.D. But
Christianity and its art did not fade overnight. Even in Egypt, Coptic art
continued to c. A.D. 1200, when Christian Nubia was overrun, although
Christian backwaters survived in Nubia until A.D. 1500 and perhaps later. And
Islam never overwhelmed the Ethiopian highlands. There Christian arts
survive to our

own

day.

Although Islamic art in northern Africa is divided into periods and given
dynastic labels, the changes in dynasties do not always correspond to changes
in style. Dynasties by themselves do not shape the arts, but the variations in the
extent of the territories ruled led to variation in artistic influences. Also,
particular religious motivations were often adduced as legitimation for new
dynasties, and exerted some influence on the arts.
Architecture was dominant in the arts of Islam. The congregational
mosque found an early 'classical' expression in Tunisia during the ninth

century. This type of building aho led to a distinct Hispano-Maghrebi

tradition that was to last until the twe ntieth century. From 1 169 onwards with
the conquest of Egypt by Salah al-Din (Saladin), Egypt developed
characteristics, often labelled Mamluk art, which survived until about 1800.
Meanwhile, in Libya and in the Sahara, a distinct tradition developed from the
'classical'

congregational

mosque.

Finally,

from the sixteenth century

took root in Algeria,

onwards, Ottoman traditions, creating new
since
1517, resisted
Istanbul
by
dominated
Egypt,
Libya.
Tunisia and
Ottoman influences. When Egypt adopted them wholeheartedly, they were
already competing with European influences (Margais 1954; Creswell 1952-9).
local styles,

THE REGIONAL TRADITIONS

Only three major upheavals affected almost all of Africa south of the Sahara
and that to different degrees. The first was the huge expansion of Bantu
languages throughout the southern third of the continent from the Cross River
basin on the border between today's Nigeria and Cameroon to southern
Africa. The expansion began some time before 500 B.C. and probably reached
the eastern Cape area in the first centuries A.D. This was not a single migration
of a culturally superior people, implanting a Bantu culture everywhere. There
was a
is no Bantu culture and there are no Bantu arts (Vansina 1979/1980)! It
common
substantial
a
recognize
can
diffusion of languages only. At best one
heritage among Bantu speakers of southeastern Africa and portions ot East
Africa, where Bantu speakers seem to have carried the practices of agriculture,
husbandry, metallurgy and settled life. But even that is not established beyond
doubt.
10

INTRODUCTION
third upheavals were the work of Europeans. They had
Africa and began trading between c. 1440 and 1500. In
of
rounded the coasts
the slave trade became dominant over all other
century,
seventeenth
the
commerce, and from the 1650s loc. 1850 a massive transplantation of Africans
arts evolved out of this
to the New World ensued. Various Afro- American
affect the arts in
directly
not
did
impact
European
this
But
great migration.
the
most of Africa to any great degree. The third upheaval was caused by
comment
the
into
flooded
that
industrial revolution and the cheap products
the
The increased European presence culminated

The second and

from 1800 onwards.

the classical regional

colonial partition of the 1880s which led to the demise of
formal partition, increased
arts in most parts of the continent. But even before
of the continent.
parts
many
and varied trade had affected the arts in

WEST AFRICA
gold trade
A.D. 750 at the latest. West Africa's past was dominated by the
The earliest
across the Sahara, and by the development of kingdoms and cities.
The largest and best
state of the Sahel may date from the first centuries A.D.
and then Mali
onwards,
A.D.
500
perhaps
known empires were Ghana, from

By

cities
and Songhay, which collapsed in 1591 (Levtzion 1973). Further east,
cities, royal
while
A.D.
1000,
perhaps
from
Nigeria
northern
in
arose
and later.
residences and states are attested in southern Nigeria from A.D. 800
date or
same
the
from
flourishing
was
Around Lake Chad, the state of Kanem
art of
well-known
a
sculpture,
Nigeria,
southern
from
earlier. But apart
West Africa, does not seem to have been focused around royal courts. Very
survived; rather, more remains
little evidence of the northern imperial arts has
focused around the city of
probably
Niger,
of the ceramics of the upper Middle

Jenne (1100-1600).

With the arrival of the Europeans, the lands on the seaboard came into the
nations on
forefront of West African trade, rather than being the last trading
by
finished
was
forest
the
of
settlement
gradual
the road to North Africa. The
in
1450 at the latest, and various forms of societies based on associations (e.g.
larger
Liberia) or chiefdoms were flourishing. Gradually some evolved into
kingdoms such as Asante and Dahomey in the eighteenth century. Meanwhile
empires over a large
in the Sahel smaller successor states had replaced the
had arisen. Early
forms
state
new
Volta
Upper
in
and
territory,
their
of
portion
association lodges, all
on, the cities, the courts, the trading settlements, the
became centres of artistic production, since masterpieces from the flfteenth
have survived. The earliest known ceramic sculpture in the

century onwards
(Ke) and
east of the area, in the delta, dates from the onset of our millennium
from the Cross River since the sixteenth century.
The evolution of the arts in West Africa never led to a massive tradition
they
emerging and covering most of the area, mainly because the institutions
abhorred
were associated with had no universal aspirations. Islam
to
representational sculpture, while trading institutions, though contributing
Africa
West
own.
their
of
art
an
with
up
some spread, were not inherently tied
influenced by northern Africa, especially in metal technologies of

was strongly

and for some decorative designs. West African arts also

again, all proof is
exercised, we believe, some influence on North Africa, but

casting, in textiles

lacking so

far.

ART HISTORY

IN AFRICA

INTRODUCTION

55
.^

g C O

>.

I*

t-

5
a Q
? rr =
a o a o jc 3 ,5
DQ DQ

QQ

Ul U,

u
13

ART HISTORY

IN AFRICA

CENTRAL AFRICA
Central African history

environments.

falls

into several sharply divided segments according to

The northern savanna saw

the rise of states near its Saharan

and the development of architecture as a dominant art certainly after
1500 and perhaps earlier. Islamic influences were, by then, already strong in

fringe

the states. Further south, people lived scattered over the land and, except for

portions in what

is

now Cameroon, no

larger political units arose.

With

the

exception of the lower Shari region, no known early works of art have survived,
barring a few pieces of sculpture from the nineteenth century.
In the Central African forest, sculpture was practised in small
communities where the foci of art were the public men's houses, centres of
government, the village temples in Gabon that fulfilled the same role, the
paraphernalia of associations and in parts the shelters where memorials for the
ancestr>' of local leadership were built. In almost the whole area, such
sculpture exhibits common traits, such as the economy of means to render
expression. Here, too, no works of art older than the nineteenth century have
survived, although some known works (e.g. in ivory) may be much older and
the traditions certainly were at least a few centuries old. On the southern
margins of the forest grew a major kingdom labelled the Kuba kingdom.
Distinctive and rich traditions in architecture, sculpture, textiles and
decoration developed there from the seventeenth centurv onwards (Cornet
1972, 1975, 1978; Neyt 1981; Perrois 1979).
South of the forest, eastern Central Africa has left us a few traces of wood

sculpture (Van Noten 1972, D. Clark personal communication) and also

ceramics from the last centuries of the first millennium. From A.D. 900
onwards, trade in copper products became important and indirect connections
with Indian Ocean trade existed. In Shaba (Zaire) chiefdoms, then larger
states, gradually grew. They flourished in the eighteenth century; some, such
as the Lunda empire, grew in direct relation to the slave trade. The Lunda
empire and the main Luba state deeply influenced all local cultures around
them from 1700 onwards at the latest. They also exercised mutual influences

on one another. This is clearly shown by the positions of central Angolan and
northern Shaban foci of sculpture. Nevertheless, some local traditions
remained unaffected, especially in the area between the Kasai and Kwango
rivers, and in the Kasai and Maniema (Zaire) areas north of the Luba empire.
The earliest sculptures in wood from Angola date from the second half of the
first millennium. Near the coast, sculpture, decorative design and textiles were
flourishing before 1500 in the kingdom of Kongo and adjoining regions. With
the arrival of Europeans, especially missionaries. Christian European art forms
dominated among the elite of that state until c. 1700 when local traditions
assimilated them. Further south, central Angolan art assimilated many
borrowings from Portuguese art.
Even in the southern savanna one cannot speak of a massive art tradition
for

much

the

same reasons as in West Africa. Nevertheless, three traditions of

and Angola, central Angola, and northern Shaba, clearly

art only, coastal Zaire

dominated the southern savanna.

14

INTRODUCTION

ART HISTORY

IN AFRICA

^'^-.

SUDAN

p^^

INTRODUCTION

EAST AFRICA

AND SOUTHERN AFRICA

East Africa is the geographic area of the continent where art traditions have
been the least studied and were the most varied (Hartwig 1978; Holy 1967).
Apart from the coastal architecture influenced by Islam, architecture was a
major art in a set of kingdoms in the Great Lakes area that had developed by the
sixteenth century, although earlier architectural traces have been found at
Bigo. In the truly drought-prone lands of southern Sudan, northern Uganda,
portions of Kenya and Tanzania, the main expression of visual art was painting
as shields, and on human bodies. Funerary
Ethiopia, nearby portions of Sudan, near
southern
in
occurred
pole sculpture
the Kenyan coast and in parts of Madagascar. In much of Tanzania the
situation is less clear. Some sculpture in clay and on wood was centred near the
Rovuma River extending from northern Mozambique as far as Malawi. None
of the arts of inland East Africa is dated and most remains are probably of
recent origin, with the exception only of rock paintings and a few ceramics
from Uganda and Bigo. But we do not even know when hunters, herders or

on rocks, on portable objects such

farmers stopped painting on rocks.

East Africa's history has been one of coastal trading settlements since A.D.
700 and even earlier, of sizeable migrations by herders, herder/farmers, or
farmers in much of the drought-prone area, of the rise of higher densities of
population and states only in the interlacustrine area and of trade routes at first
only from Malawi through southern Tanzania to the coast and then, in the late
eighteenth but especially in the nineteenth century, from the coast to beyond

Lake Tanganyika, the Great Lakes and northern Kenya, led by Muslim
Swahili speakers from the coast. This corresponds well with what we know
about the practice of the arts, which is clearly far too little.
History south of the Zambezi shows a development of agriculture and
later specialized herding as well as the opening up of gold mines and the export
of gold to the Indian Ocean by A.D. 1000. Thereafter, centres characterized by
both architecture and sculpture arose first at Mapungubwe in the Limpopo
valley, then in Zimbabwe, where centralized regimes dominated the area until
well after Portuguese settlement on the coast.
From Zambia to Transvaal, small-sized ceramic sculptures are found
from c. 1000 onwards. Earlier still, ceramic sculpture had flourished in Natal
and southern Transvaal between c. A.D. 500 and 800 (Maggs and Davison
1981). In later times, well-watered southeastern Africa developed a mixed
economy of herding and farming. Groups there began to form various
chiefdoms, until the nineteenth century when a Zulu leader, Shaka, built a
large army and a large state and plunged the whole region into turmoil. New
formations of large size appeared as far north as western Zambia. This
upheaval ended a long sequence of artistic evolutions in Zimbabwe, Transvaal,
and portions of the Orange Free State that had been characterized for many

political

centuries by architecture in stone. The great miseries in the area from the
1830s onwards resulting from the Zulu upheaval and from the Boer migration
into Transvaal, probably brought

The western

most

artistic traditions to

half of southern Africa

is

an end.

desert or at least very drought

prone. Since the end of the first millennium, only pastoral nomads shared
these lands with hunters. The rock art of South Africa and Namibia from

17

ART HISTORY IN AFRICA

LESOTHO
SWAZILAND

Tswana

Zulu

400

800

Km

Fig. 1.8 Southern Africa

than 27 000 years ago is the work of these hunters (Inskeep). The
tradition died out in the late nineteenth century as they were decimated by
Bantu speakers and Europeans alike. The arrival of the Europeans on the Cape
and the Dutch settlement there from 1652 onwards brought European styles
into the area. But they had little effect on the arts beyond the areas of their
settlements. In fact, European art influenced the African arts less here than on
the West African coast. Although there was some expansion from the Cape
settlers before the 1830s, an emigration of Dutch speakers, the Boers,
dissatisfied with British rule, occurred in that and the succeeding decades.
This led to the expansion of European settlement ultimately all over southern

earlier

Wherever they settled, African classical art died out.

Madagascar was settled by immigrants from southeast Asia and from the
nearby African shores before A.D. 1000. As nodes of population slowly grew on

Africa.

the island, larger political units gradually arose.

18

A kingdom

of the Sakalava

INTRODUCTION
dominated in the seventeenth and eighteenth centuries, while a state of the
Merina overtook it from the late eighteenth century onwards. By the 1890s the
Merina had control of more than two thirds of the whole island. The arts of the
island have always centred around tombs. Some Islamic tombs of East African
parentage are found in the northwest and may be older than 1500. Elsewhere
the architecture of tombs, their sculpture and their decoration is wholly
original, but the age of this tradition is not well known. As towns developed
architecture
first Majunga on the West Coast, later Tananarive in the centre
became more important. In the nineteenth century a Malagasy architectural
style became prominent. It was heavily influenced by European models as well
as by older methods of construction in planks. The affiliations of Malagasy art
and architecture are not at all clear. Indian, Melanesian, Arabian and
Indonesian influences have all been cited, and an African influence from the

Mozambique

coast

is at least likely.

THE STUDY OF AFRICAN ART

first arrived in Europe as curiosities for
princely cabinets, the oldest mention of a work of art being one of 1470. Islamic
art had been known and had exercised some influence on European an long
before then. It is not surprising that an awareness of this art developed as the

Art objects from sub-Saharan Africa

history of art

was emerging

as a discipline,

although the main scholars

who

studied art were active only after 1900.

By the late nineteenth century, objects from sub-Saharan Africa were
acquired and housed by European museum curators as specimens of material
culture. Concurrently, curators developed the first methods of cataloguing and

The looting of Benin by a British expedition in 1897 led to

the realization that there was art in Africa. Benin's naturalistic art work in
labelling artifacts.

'bronze' and ivory appealed to

most

royal collections in

European

Europe.

An

tastes.

Some art work found its way to

word 'art' in connection

early use of the

with Africa appeared in the title of A. Pitt-Rivers' descriptive book of 1900,

Ancient Works of Art From Benin. Only L. Frobenius, later a famous culture
historian of Africa, had used the term earlier (Frobenius 1896). The Benin
boom led to the first art historical study, F. Von Luschan's massive Benin
Antiquities (3 volumes in the original!) of 1919.
A wider recognition of African arts began in 1905 when European
avant-garde artists in France and then in Germany recognized them as such. A
wave of almost delirious enthusiasm followed and gave rise to the first private
collections and to C. Einstein's Negerplastik (Leipzig 1915), which set the tone
for a spate of lyrical works that followed. Only artistic form mattered, social
context and meaning were irrelevant!
Anthropologists began to redress the balance after 1925. The first field
trip specifically directed towards an examination of art dates from 1933
(Bernatzik 1933; Gerbrands 1937:52-65). By 1945 a specialization in 'African
art' was beginning to emerge in anthropology. Given the characteristics of the

was unsustained.
Then, from 1956 onwards, archaeologists who viewed African art seriously,
discipline at that time, interest in the history of the arts

entered the

field.

The

historical perspectives

developed only very slowly until

19

ART HISTORY IN AFRICA

Now

commitment to the study of the

growing.
These developments have been immeasurably affected by the
concomitant growth of a market for African art. Even today scholarly studies
have insufficiently disengaged themselves from mercantile interests and are
partially crippled by them. Meanwhile, for over a half century, large private
collections have been assembled, and have saved many art works from
destruction. Nevertheless, the proliferation of forgeries, the keeping of precise
origins as trade secrets^, and the publications, written or directed by dealers,

recently.

the pace has quickened and a

history of art in Africa

is

touting their holdings, have certainly blemished the services that the market

has rendered.

For much too long an attitude fostered by some entrepreneurs of the

market place was that of African art objects aiS art trouve, 'found art', like pretty
pebbles on a beach, and further fed the aesthetic involvement of collectors by
titbits of exotic sensationahsm. Unfortunately, students of African art were not
totally free from such influences and still must beware of them. And so it came
to pass that 'African art' was the sculpture of West and Central Africa, and
that, until recently, dating it did not matter. Now, however, there are changes
in the market; induced perhaps by archaeological successes and the passage of
time. Now the greater age of an art object enhances its value and an attribution
of date becomes a matter for expertise of authenticity.

20

CHAPTER TWO

IDENTIFICATION
LABELS, OBJECTS

AND DOCUMENTS*

visitors to museums or readers often just glance at an object or its

reproduction and then look for a label or caption such as 'Ekoi headpiece' or
'Qait bai mosque, Cairo' before they turn to a study of the work. Labels and
captions provide identification. They are coordinates in space and time and
allow the 'placement' of objects in relation to others, because they form a grid
according to which any work of art can be ordered.
Minimal labelling should consist of three items of information: a reference
to the object, the name of the artist and a date. Very often, for African art, the
name is replaced by an ethnic name and the date is omitted simply due to lack

Most

of information or to lack of interest in a date. Omission of the date indicates

how much art history in Africa, especially south of the Sahara, still remains in
its

infancy.

Labels are felt to be essential by most viewers. The art work by itself does
not give this information. It must come from other sources. Thus, looking at
the Ardra board (Plate 1.1) does not tell us that it came from Ardra in the
seventeenth century. We should also know that it is now at Ulm and that
seventeenth-century catalogue data exist (Willett 1971:81-2). All of this is
documented in writing. And because documentation helps provide

and time, documents play a central role in art history.

between object and document is rarely foolproof.

identification in space

But

the

link

Inscriptions on monuments provide the closest relationship possible, at least if

they are contemporary with the making of the object. If so, the link can hardly
be faulted. A piece of Fatimid pottery signed 'Sa'ad' (Talbot Rice 1965:92)
refers to that pottery, no other. A dated inscribed frieze in a mosque can be a
direct indication of the date that mosque was built. In all other cases,
refer to other objects than the ones they now are associated
sixteenth-century text about a salt cellar in ivory from African cannot
automatically be taken to refer to one of the known salt cellars with African

documents may
with.

features.

The description of objects in such documents becomes quite important.

Photography, the perfect description, now used by most museums as part of
catalogue records, is a practice no older than a generation or two. Formerly
either a drawing or a detailed written description gave the best guarantee that
one item would not be mistaken for another. But such notices and drawings
were rare. The usual mention of objects was found on ledgers, catalogues or
*

F. Grathner,

Meihode der Eihnologie, Heidelberg, 1911, pp. 12-54.

21

ART HISTORY

IN AFRICA

is, on lists. Lists do not usually describe items in more than a

few words and do not systematically include drawings.
Hence any critique of identification starting from documents must begin
by examining the possibility of error. As there are two sources involved - the
object itself and the document - they first must be confronted one with the
other. Examination of medium, technique and style (and theme to a lesser
extent) tends to group objects in sets of similar features. If the documentary

inventories, that

attribution runs contrary to these features,

labelled 'southern Zaire'

it

on the museum record

is
is

suspect.

An

Isis statuette

either evidence of a hitherto

unsuspected spread of late small old Egyptian statuary to Zaire, or the labelling
is wrong. In this case, the error in the label could be established, as the
collector in Zaire had stopped over in Cairo on the way home.
The procedure of giving preference to identification by the features of the
object itself over its attribution by record can lead to serious error. The typical
case is an attribution by document that is not confirmed by an attribution of
technique or style, usually because these features in the object are not found on
the corpus of known works from the area and of the time. Either the object is a
forgery or the document is in error. But the characteristics of 'typical' works
are developed out of an examination of an extant corpus and - especially in
Africa - there is no guarantee that the corpus used is in fact representative.
Thus, an 'atypical' object may be genuine and the attribution correct. A
hitherto unknown variant has then come to light. The weakness in the
argument to reject the documentary evidence is that it is based on negative
evidence, that is, the absence of 'typical' features in the object. In order to
stick, the extant corpus must be totally representative, a difficult proof to
provide, when innovation is a recurring part of the creative process.
Discrepancies between motif and the object itself for one feature should
be tested against others. Is the wood of the carving similar to the wood of the
corpus? Is the technique similar? Is the style comparable? Is the theme
rendered in the corpus? In the case of fakes or errors several characteristics
differ and multiple negative evidence accumulates. Doubt increases.
The proof of the pudding comes when the characteristics of the object
quite clearly refer to a body of art works different in place and time to the one
referred to by the label. For now there is positive evidence. If in addition we
can show how the documentary error could have arisen, the case is clinched.
Thus, the so-called Vallisnieri pieces at the Pigorini museum in Rome were
said to be Chinese (Plate 2. 1). On the base the legend states: 'Idolo de la China'.
Further attributions on the bases eventually allowed the pieces to be traced to
Central Africa, to the Kina district of the kingdom of Kongo, which confirms
the comparison of the objects with other carvings. Detective work among
documents not only established in the end where the statuettes came from but
how they reached the Pigorini museum (Bassani 1978; Bontinck 1979). The
Capuchin missionary, F. da CoUevecchio, probably acquired them between
1690 and the close of 1694 in Kina, brought them to Lisbon in 1695, and gave
them to the nunzio G. Cornari, who carried them to Padua in 1697. Upon his
death, A. Vallisnieri acquired them in 1722 and from his collection they went
to that of the University of Padua in 1730 from where they were ceded to the
Pigorini museum in 1877. The confusion of Kina and China is explained by the

22

IDENTIFICATION

Plate 2. 1 Female Figure. Wood. Use: unknown. Kina (old kingdom Kongo).

Height 24-8cm. Before 1694.

Known

as Vallisnieri Figure.

One of a

set

Museo preistorica ed einografko, Rome.

of two

23

ART HISTORY

IN AFRICA

identical spelling used in seventeenth-century Italy.

The case above shows clearly how important documents can be. For older
moveable objects usually have gone through a history of ownership and this
can only be traced through documents, mostly descriptive catalogues, but also
other acquisition data. Moreover, documents are even more important for an
understanding of the social and cultural background of objects. The mention
of desecration in the entry about the Ardra board, for instance, is important.
They are nearly as important as the art works themselves. And they should be
examined by the ordinary rules of historical evidence, for very often they are
vitiated in one or other respect. The catalogue entry for the Ardra board may,
for instance, have been written from memory many years after the acquisition
and may inaccurately stress some of the sensational details adduced then.
In the following sections the identification of art objects will examine a
sequence of questions to be asked: What is the work? Is it authentic or is it a
forgery? Who made it and where? When was it made? For if we have answers
we have established the position of a work of art
comparative grid provided by space and time.

to these four sets of questions,

in the

THE WORK

ITSELF: DESCRIPTION

The first question to ask is: what is a thing and how was it made? Usually there
is little difficulty. The Ardra object is a flat board. Common utensils and tools
are recognizable as such. But mistakes can occur. What looks like a stool could
be a throne and a jewel could be a weight. The first danger for the onlooker is to
go beyond the shape of an object and assert a use for it that may be wrong.
Exotic looking knives from the Zaire River are knives, not 'executioners'
weapons'! The examination of an object must remain factual. Questions of use
are to be dissociated from questions of appearance; these will be examined
later. Nevertheless, speculating about use helps in the definition of an object.
If we suppose an object to be a mask, we will expect slits for the eyes, edges
rounded by use, remnants of cosmetics or perspiration on the inside, ways to
slip it over the head or ways to fasten it for carrying. We will wonder about its
weight: can it be carried by a dancer or not? If we suspect a ruin to be a
mosque, we will be careful to record the orientation, and search for evidence of
a mihrab, the niche in one of the walls that always should be oriented towards
Mecca. So even though an examination of use comes later, suppositions about
such
it help in the physical description of the object by drawing attention to
details.
It is

not enough to describe the object as

it is

seen.

The description should

medium. The Ardra board we used as an example appears to be

made of wood. But what kind of wood? We do not know; no one analyzed it.
And unfortunately that is the usual situation in African art - cursory inspection
is deemed to be enough, but it is not. Only laboratory analysis will settle such
matters. No West African copp)er-like objects had been analyzed before the
1960s (Shaw 1970a, 1978:182^). So-called 'bronzes' are now seen to be

establish the

copper, brass, leaded bronze,

less

leaded bronze, etc. These raise questions of

technologies, attributions to identical or different workshops, problems ot

supplies of the raw materials and so on. Similarly, iron can be high carbon steel
or pig iron. Bricks are never just bricks.

24

They come in standard

sizes that differ

IDENTIFICATION
from age to age and from area to area. The bricks of the great mosque at
Qairawan are of a standard of dimensions used in ninth-century Tunisia ruled
by the Aghlabids and thus confirm other indications about the age of the
mosque (Margais 1954:42). Wood can be very diverse. To rely on what people
say the wood usually used for this or that item is, will not do. The answer is
often guesswork and involves the relative prestige of different woods in a
society, whereas a laboratory analysis settles the matter (Dechamps 1970/8).
The results of analysis, in turn, give insight into which kinds of woods were
used (and the number is always restricted), what they were used for, and
sometimes we can even fmd out why such woods were used. The profile in
wood use of one community at one time differs from that of other communities
and may vary over time.
The description of an object must also include the traces left by the
technology used. In buildings, walls can be of brick, bonded or not by one or
another type of mortar, and laid in different systems of courses. Walls can be
rubble inside, dressed stone outside or stone and wood. A metal object may
have been hammered or cast and traces of the casting process used can usually
be found. Painting techniques differ and traces of the media and application
are quite visible. Patina on wood can be analyzed and the traces of carving by
adze or knife or axe are evident as are the traces of polishing by metal file as
opposed to sanding by leaves or other materials.
Once the description of the object as it now exists is done, there comes the
more difficult examination of whether that was its original appearance, and
whether it contains hidden features. The latter is the easiest to discover. When
peculiarities in the mass of objects are suspected. X-rays will reveal them. It is
not infrequent for sculpture from West or Central Africa to hide features
inside. Swords from the Kuba kingdom (Kasai, Zaire) often bury a tiny
olivancillana nana shell in the hilt. Beneath a rather shapeless exterior some
western Sudanic (Senufo) sculptures contain a well-carved statuette and some
objects from shrines of western Nigeria (Yoruba) contain paired statuettes in
metal Hnked together (Claerhout 1978; Herreman 1978; Neyt 1981:87-9).
Such cases tell us that the objects are not what they seem to be to the Western
eye. Explanations for such practices must be sought in the cultures of origin.
The most notorious problems of original shape as opposed to the shape
now perceived are found in architecture. The great mosque of Qairawan was
the most revered in the Maghrib. It was remodelled many times. As we see it
now, the minaret is its oldest part and may date from c. A.D. 724/728. The basic
ground plan also dates from that time or even earlier. The main prayerhall
dates from 836 with extensions of 862. Remodelling, and the addition of
courtyard galleries, occurred in 1025, and further additions date from 1293 to
1316 and even later. Restorations occurred in every century after that. The
mosque has a history of shapes that must be unravelled. As the most
prestigious mosque in the Maghrib, it was an object on which successive rulers
in Tunisia liked to leave their mark, to aggrandize it or to restore. Additions are
detected by a consideration of the present ground plan, the tie-ins of different
parts of the masonry, and examination of the different styles of execution,
epigraphic evidence and documentary reference. Even in this well-studied
still not completely agreed about the sequence of
construction (Margais 1954:9-22; Lezine 1966:12-81; Sebag 1965).

case, art historians are

25

ART HISTORY

IN AFRICA

is not alone in this; a number of carvings or paintings

original shape as decay or erosion takes its toll. More
their
not
retained
have
importantly, many carvings, for example all masks, were only part ot a total
shape that included feathers, beards, costumes, paint, etc.; the carved parts

But architecture

have been selected out of the whole. Objects should also always be examined
with the possibility of restoration or mutilation borne in mind. 'Total
restoration' means totally new fabrication, and requires a search for the object
which the restorer copied. Moreover, many of the objects that reach collectors

and museums are

FORGERIES

fakes.

AND THEIR ILK

objects are forged for all sorts of reasons, but by far the main
motive is gain. The more flourishing a market in art, the more fakes. Old
Egyptian objects were already forged in the last century, while Benin fakes
appeared on the European market less than three years after the looting of
treasures from that city (Graebner 1911:12-21). Shortly after 1900 a Fang
knife (Gabon) with a cuneiform inscription appeared on the market. The knite
was authentic, the inscription not (Graebner 1911:18). Fakes have multiplied
since then - some are made in Europe, some in Africa by African artists
copying older works from the same or from a different area. All of these are
forgeries in the sense that they do not correspond to the indications of origin
attributed to them (place and date) when offered for sale. Quite apart from
out-and-out imitations, some art dealers forge pieces by restoration or, much
worse, by destroying genuine pieces o use the parts obtained for several new

Documents or

Between this and disfigurement of authentic but stolen pieces, the

boundary becomes very thin. An extreme case is that of the Chibinda Ilunga
hunter's figure of the museum of Belem which was stolen and then mutilated
by the removal of small figures on the base (Bastin 1976). Many private
collections as well as even major museums harbour fakes and many are
unwittingly reproduced in art books. While the owners and curators do not
talk much about it, still, some pieces are only exposed as fakes many years after
pieces.

their acquisition.

Careful study of a piece frequently exposes forgery. Examination of the

study of the
techniques often yields traces of European tools and signs of haste when the

medium can uncover unusual features or even a foreign substance;

forger has tried to use techniques identical to the original. A forgery crafted
with loving care would not afford much of a profit, at least not until recently.
The finish often betrays the forger: polishes and patinas are easily analyzed and

hard to imitate, artificial ageing by induced corrosion, addition of mud,

smoked crusts, even fake insect boreholes, can also be detected rather easily.
One should not stop at the examination of one feature, but continue
throughout the whole set that can be examined. Style by itself is a lesser
indicator because this is the characteristic forgers imitate most. As to theme
and motifs, most fakes are faithful copies of some original (which often can be

found to be

a reproduction in a

well-known magazine), but some of the best

innovative in theme in order to escape detailed stylistic

comparison and to attract a higher price as an unusual or a 'unique' piece.
If serious examination can nearly always lead to the exposition of

forgeries

26

are

IDENTIFICATION

why then do museums and collectors not undertake them routinely?

Mostly it is a matter of cost, and legal authority (pieces not yet acquired cannot
be physically analyzed). Whatever the reasons, there is unfortunately too
much laxhy over acquisitions. Even in 1980 we saw a major museum acquire a
large art work representing a hitherto unknown theme for the area it was
supposed to stem from, in an unknown type of composition and accompanied

forgeries,

by very slender documentation, without analyzing the wood, the patina, or

even attempting to account for several unique and foreign-looking motifs on
the piece. Pressure from donors was cited as a reason for the acquisition. The
only step taken by the management was to seek certificates of authenticity by
consultation with authorities in the field, none of whom actually said in so
many words that they believed it to be a forgery. But then, they did not
guarantee the authenticity either.
If forgeries occur most frequently among moveable objects for sale, a few
have also been reported for Saharan rock paintings (Lajoux), where the motif
was a hoax, of a type sometimes played on unwitting archaeologists by
members of the public carrying Roman coins (Mauny 1970:79-80).
Documents faked to substantiate information about pieces seem much rarer
than forgery of objects themselves, and forgery of ethnographic information in
is exceedingly rare, although not wholly unknown (Piskaty 1957).
Supporting documentation and unmoveable objects of art should then also be
scrutinized carefully for any signs of tampering or forgery.

general

AUTHOR AND PLACE OF ORIGIN

For any object to be more than an objet trouve, a 'found object' that we can
admire but never understand, identification of origin is the first step. For
public monuments, wall frescoes and other immovable objets d'art, the
question of provenance is straightforward. For movable objects, it can become
quite a search as the previously mentioned Vallisnieri statues illustrate. For
most of such objects, the indication of origin is reported at the date of
acquishion, at least in a public collection. Mention of the name of the artist was
remains rare today. Most artists, except for
and
very few works were ever signed; collectors in
architects, were anonymous
the field merely noted the place where they had gathered the piece.
Unfortunately, even that information is often unclear. At the beginning of the
century, the mention of a district or a market town was often enough. Even
then, and probably under the influence of the earliest ethnographic museums,
the practice was to label a piece by a tribal origin, that is, by the name of the
ethnic group where it had been found. Since then, such ethnic names have

exceedingly rare and

still

main nomenclature, although some indications of village of origin

in the 1930s and later, but this trend was reversed by the late
appear
began to
1950s. By this time major dealers in the arts of Africa south of the Sahara could
send out personnel to buy up or acquire the production of any locality where
classical art was still to be found, and scholars, seeing that their lists of villages
of origin merely provided guides to buyers, stopped providing such
information. The very existence of a flourishing and aggressive market has

become

the

thus been quhe detrimental to the identification of art objects.

Many objects came to Europe or America without any documentation,

27

ART HISTORY

IN AFRICA

IDENTIFICATION
and

until recently there

was a rubric in African Arts, asking for identification of

reproduced there. Identification in such cases is provided by

comparison to objects whose provenance is known, usually by overall
similarity of the icon and by detailed stylistic comparison. Specialists become
pieces. Thus,
so versed in this that they almost unconsciously identify
even if we did not know that the ifa board of Plate 1.1 came from Ardra, we
would still place it in western Nigeria (Yoruba) because ifa boards from the
objects

nineteenth and twentieth centuries (cf Plate 2.3) are still quite similar in
general shape and style. Some over-confident art critics identify workshops in
this way, and in some cases artists. 'This work betrays the same hand as the
master of the aquiline profile' means that the writer claims that both pieces
were made by the same artist. It is also a statement that is usually impossible to
check because the maker(s) of any of the works compared remains quite
as indicating stylistic groups and
but not providing absolute
manufacture,
presumably close relationship of

unknown. Such statements should be taken

identities.

Even when

places are given

carefully. Quite apart

from errors

by records, they need

to

be examined

as in the Vallisnieri case, the locality in

Africa may be that of acquisition, but not of manufacture. Thus, the noted
German anthropologist, Frobenius, used to send out helpers to collect objects.

Bolombo, merely means that it is in that locality on the

Sankuru that he acquired the object. His collector might have obtained it up to
well over fifty kilometres or so away in a circle all around. In another case a
mask was acquired in Tabora, Tanzania before 1898. Yet its style is clearly
central-east Luba (see Fig. 2.1). Very likely it came to Tabora along the
caravan routes (Frobenius 1898: Fig. 12, Tafel V; Krieger and Kutscher
1960:84, Bild 73, n. 161 and Bild 75, n. 165). As trade and sales of objects were

An item of his,

not

labelled

Africa even south of the Sahara, the latter situation

at all rare in precolonial

may have been much more common,

especially for early objects, than

is

realized. We know, for instance, that the kings of Mbailundu in Angola had
their thrones made to order further east among a people called the Chokwe
north
(Bastin 1968/9:60, ill. 43), and that carvers from the Loango coast, just
the
as
north
of the River Zaire, operated as far south as Luanda and as far
de
Fleuriot
1904:308-11;
(Nassau
century
Ogowe river delta in the nineteenth

Langle 1876:294-5). In

all

such cases

stylistic identification will

usually

correct the written entries.

on the ingrained European belief that

from all others in its customs and
differed
group
each then-catalogued ethnic
especially, apparently, in its visual arts, while all the members of the 'tribe', on
the other hand, wrought art objects in the same style. But for more than a
generation now, scholars have shown that tribalism is a colonial phenomenon,
that ethnic feelings are variable and ethnic identities multiple, overlapping,
and generally fuzzy. Nevertheless, this remains the main nomenclature today,
and it does lead to inaccuracies and muddles. As a nomenclature the system is
sometimes absurd. A huge area in the west African savanna is labelled Bamana
(or Bambara), a huge area in the Gabon/Cameroun forest is Fang, but every
single town in the Cameroun grasslands is an ethnic unit, such as Bamessing,

The

identification

by

'tribe' rested

Fontem (Bangwa), etc. We know from experience that several styles are
be found in the one Bamana area but we still label it Bamana. Or we have a

Bafut,
to

29

ART HISTORY

IN AFRICA

Plate 2.3 Divination board.

Wood. Used

in

ifa

divination.

Northern Yoruba, western Nigeria.

Museum

Volkerkunde Berlin-Dahlem. Width 40 5 cm. Before 1910. The face represents the God EshulLegba. Compare
knots with Fig. 9.4. Probably North African element introduced via decorative pattern on costumes

30

fur

interlace

IDENTIFICATION

Fig. 2.1

Luha mask. Acquired at Tahora, Tanzania, from' Manyema' people in 1890, showing how far objects could be

spread by traders. After L. Frobenius 1899, Tafel 2.12a

label

Igbo for eastern Nigeria, where the major characteristic is that style varies
from one subgroup to another. In Liberia workshops are known that

greatly

Wee (Gere) manner or in the Dan (Gio) manner, as the

and these four labels refer to two peoples divided by the
Liberian-Ivory Coast boundary (Himmelheber 1960:13647; Fischer 1978).
Contrary to Fagg (1965), there is no necessary link between ethnicity and any
form of visual art. True, in many cases features of body decoration, such as
scarifications or hairstyles, or features of costume such as distinctive caps or
textiles, become badges of identity, but other expressions of visual art rarely
become a focus of ethnic pride. Anything can, on occasion, become such a

can work either in the

client wishes,

language in the Cameroun grasslands; a certain ritual of female

among the Tio (Brazzaville); a form of greeting among the Mongo
(Zaire bend); even forms of marriage, so why not on occasions a type of object?
Because actual occurrences of such situations seem to be so rare, each case
requires proof. The focus of identity should lie on the object, not on the
institution of which it is a part, and it should lie on a major formal feature, not
on a detail, such as facial scarification that identifies the ethnic group of a
statue, for instance. Then one can reasonably apply an ethnic label. The
exceptional nature of such cases does not warrant a convention whereby all art
is labelled by ethnic group. Why not label it by institution?
An example from Gabon shows how serious distortion can result from
such a perspective. A type of reliquary figure, largely two dimensional, made
out of wood covered by copper and/or brass strips or wire, was made in eastern
Gabon and is known to collectors as mbulu ngulu, not quite an accurate name.
These are 'Bakota ancestor figures' (see Plate 5.1). A recent study of the
different substyles asserts once more that these are Bakota pieces. But the
focus:

possession

31

ART HISTORY

IN AFRICA

SANGO

\
"^

^
\

NZABI

^-

mbamweX

twUMBUJ
'^'s.

Key

TSANGI

^^^s^

IDENTIFICATION
There is
no relationship between language and the manufacture of reliquaries.
The art historians first equated language and culture, then equated culture and
aesthetic creation and finally claimed that some groups which had mbulu ngulu
but were not Kota speakers must be 'Bakota' after all, while omitting the
unmistakably Bakota Akota! Having done this, they are still faced with the fact
that some Bakota use masks belonging to one tradition, some use masks of
another tradition, a third group have their own masks christened Bakota
because the Bakota proper, east of the Ivindo, have them and some Bakota may
never have made masks at all. Moreover, none of the styles in masks
corresponds to the style of the reliquaries. The facts are clear enough by
division does not correspond to a coherent area of manufacture either.
in fact

themselves, only the attempt to force them into an ethnic mould clouds the
issue (Chaffin 1979; Perrois 1979; Dupre 1980; Siroto 1981).
It is high time to abandon this artificial nomenclature (as against Fagg
1965; Bravmann 1973). Objects should be labelled by village and workshop of
origin, if known, otherwise by reference to the institution to which they are
associated. In this book I use geographic designations and put ethnonyms in
brackets, but I cannot do away with them on penalty of leaving my readers

without any points of reference when they consult the literature.

In northern Africa, among the oikoumenical traditions, identification of
common movable objects such as rugs, textiles, ceramics, works in leather or
wood is just as complicated as south of the Sahara. Often identification by style
is not as certain a guide here as it is in the regional traditions. Thus almost
identical plaster decorations were found at Sadratha and dated to c. A.D. 1000,
in southeastern Algeria and at Boudenib, dated c. 1900 in Morocco. Places of
acquisition can be very far indeed from place of manufacture, as a caravan
trade or trade by sea carried such objects as far as the East African coast or the
northern borders of Akan country in Ghana. On the other hand, monuments
are much better documented, often because of inscriptions. Especially in
Islamic art, costly textiles, metalwork and some ceramics carried the date of
manufacture and sometimes the name of the artist (Hill and Golvin 1976; Atil
1981:51, 148-50; Talbot Rice 1965:92, 148-50, 188). Inscriptions were
treasured here and practically all major mosques have older or younger
inscriptions relating to their foundation.

DATING
of dating can hardly be overstated. No history without
chronology, whether absolute or relative, is possible, and that is as true for art
history as for any other branch of the discipline. Unfortunately, very few art
historians south of the Sahara have heeded it. In northern Africa and Ethiopia
the relative occurrence of dated monuments and of literacy have insured that
chronology attracted at least some attention. South of the Sahara the problems
were greater and they were ignored. A paper first published by F. Olbrechts in
1941 (Olbrechts 1943) is still the only general discussion of the problem to
date, forty years later, the same neglect bewailed by Olbrechts prevails. Yet

The importance

several

means of dating

exist,

by physical means applied

whether by using written and

to the objects themselves,

oral

documents or

but clearly the effort

33

ART HISTORY IN AFRICA

involved seems too great for most practitioners. There will, however, be no
history of art if there is no chronology.
The written data linked to the object usually include its accession to a
collection, sometimes its acquisition in the field, but almost never the date of
manufacture. For some objects that date may be the year of acquisition, but

sometimes the work may be much older. Objects that have been collected a
long time ago, for example, the Vallisnieri pieces, have moved and must be
traced down from descriptive catalogue to catalogue and from references in
other writings. We have already seen that a major problem here may be to
establish with certainty that an object mentioned in a text is the same as the
piece we know. Hence the examination of iconographic materials is of unusual
value here.

probably of African origin in European

of Charles the Bold (Brussels, Chamber
accounts
the
in
contained
collections
of Accounts No. 1925, f. 348; April 1470): 'To Alvare de Verre, servant of Sir
when lately he
Jehan d'Aulvekerque, Portuguese knight ... 21 pounds
.'.
him.
idols
to
as
wood
in
figures
presented a sword and some
European archives, repositories and printed documents cannot usually be

The

earliest reference to objects

is

related to objects

now

extant, even in curiosity cabinets, but relate to objects

seen in Africa, or others brought to Europe and now lost, as in the case of
Charles the Bold. But as similar objects still exist, the general conditions of
their manufacture, the location and date of manufacture and the general
patterns of acquisition and removal to Europe can be known. Thus Portuguese
account books and stray references from the 1500s onwards document their
acquisition of ivory spoons, forks, salt cellars and horns in Sierra Leone and on
the coast near Benin (Ryder 1964; Fagg 1959). Some of these are

Afro-Portuguese work because some of the extant objects, as on horns, show

inscriptions in Latin and motifs borrowed from Iberian works of art. Others
are entirely uninfluenced by this. An occasional detail helps to date actual
objects. On the salt cellar now at the Pigorini museum, V. Grottanelli observed
that the men wore trousers with a codpiece fashionable in Italy around 1500
and reproduced on Luca SignoreUi's Last Judgement at Orvieto (St Brizi
Chapel) dated between 1499 and 1502. But codpieces remained f ishionable for
a long time afterwards. A horn could be dated because it commemorated the
marriage in 1500 of a Portuguese king and therefore dates from 1500 or
later, and such dates agree with the archival references.
other pieces, dating by comparison with European drawings of
costume or indeed by comparison with extant costumes themselves is possible.
In this way a number of Afro-Portuguese ivories can be dated within a decade
or two. Late rock art in South Africa can be dated when Europeans or

sometime

On

by the costume or the shape of the tools. In other

Thus, central Angolan
and later, while staffs
century
sixteenth
the
thrones imitate Iberian chairs from
of chiefs often sport copied rococo motifs, dating them by association to the

European

tools are indicated

cases relative dates

emerge

eighteenth century or

Beyond such

in similar comparisons.

later.

references, written data

and

illustrations provide general

Africa south of the Sahara, al Bakri (c. 1067/8) and

dating
Ibn Battutta {c. 1352/3) tell about sheet gold and carving in the West African
empires of Ghana and Mali. European authors begin to mention art from the
for artistic activity in

34

IDENTIFICATION

Plate 2.4 Bowl. Ivory. Used as salt cellar in Europe. Not necessarily made for export, hut previous use unknown.
Sherbro Peninsula (Bulom), Sierra Leone. Museo preistonco ed etrwgrafico, Rome. Height 43 cm. Probably early 1500s.
Restored,

and

battle

axe not authentic. European codpiece and

trousers date

it

post

1499

35

ART HISTORY

IN AFRICA

mid-fifteenth century onwards, both on the Niger bend and along the coasts
and shortly after that African 'idols' are mentioned in Europe itself. Such
written sources can be very important, for instance they tell us that curios were
made for sale to Europeans on the coasts of Angola and Gabon as early as the
1850s and that by then, and indeed earlier, craftsmen from Loango, north of
for
the mouth of the river Zaire, were working as far north as Gabon. Except
very early references, and later the best known printed works, art historians
have not searched for such sources, yet there is little doubt that written sources

should be vigorously explored. From sampling them, I believe that nineteenth

and eighteenth-century sources especially have been neglected. Many insights
into the arts of sub-Saharan Africa remain to be found there.
Writing and epigraphy date most major monuments in the oikoumene,
chronology of
barring only buildings earlier than 1600 in Ethiopia. In Nubia, a
of bishops
wall paintings has been built up by the combined use of the list
there
found
frescoes
of
found on a wall in Paras and the superimposition
miniatures
for
except
undated
remains
Ethiopia
(Michalowski). Painting in
where the associated manuscripts provide dates and for some wall paintings
which can be dated by details of costume or motif (Leroy 1967; Chojnacki
1973).

Oral data can also contribute to chronology, even though traditions are
attribution to a
to be notoriously weak in precisely this regard. The
stylistic
particular king of each of the royal Kuba statues (Zaire) agrees with a
Shaba
In
onwards.
1750
from
c.
sequence of statues and allows for a dating

known

of one master of the famous Buli style was remembered by

who kept an ancestral statue he carved for a line of
have lived in
princes. Twelve guardians succeeded each other. The carver may
in Benin
data
Oral
464).
1977:320-1,
444,
the early nineteenth century (Neyt

(Zaire), the

name

the hereditary guardians

major art forms

are often cited as hypotheses regarding the sequence in which
were introduced at the court. On the whole, however, the documentary value
ultimate
of such data remains slender. While they are very important as
sources for interpretation or data about use, they rarely deal with chronology.
When they do, they most often relate the developm.ent of this or that art form
does
and not the history of panicular pieces. Nevertheless, oral tradition
provide, from time to time, important information.

PHYSICAL DATING*
estabUshing
Dating the objects themselves mostly reUes on physical means of
and large,
By
laboratories.
of
up
setting
their age and this reqmres the
of
development
the
for
and
facilities
such
for
clamoured
have
archaeologists
Nevertheappropriate techniques, but museum personnel remained passive.
less,

dating can

now be securely done on older objects of wood or of terra cotta as

See Fleming 1977 and Iskander 1980.

Plate 2.5

Four

figures with raised hands

one block. Wood. Use and meaning unknown. Bandwgara

cliffs,

Mali.

T. Northern,
L Wunderman collectum. Height 48Scm. Carbon 14 date wah uncorrected half-life and uncaltbrated by
Dr SorthemivxU publish
whom I thank vers, much: 1305 .\.n. 90. Range .\.D. 1215-1.^95. Prohahly Bih-14thc.
^-

30

dates from similar sculpture

known

36

as

nommo

and

and

labelled

'

tables.
discuss their true age, taking into account dendrochronolog%'

Tellem'

Such figures

are

IDENTIFICATION

37

ART HISTORY

IN AFRICA

Malt.
Plate 2.6 Figure with raised hands. Wood. Use unknown, except traces ofsacnficial matter. Bandiagara cliffs,
nommo, Tellem. The
Koninklijk Insiituui voor de Tropen, Amsterdam. Height 48cm. Acquired 1973. Also known as
could be from c.
patina is derived from ash, dried animal hlood and dried millet gruel or beer. The date of manufacture
pieces, whaieverthe
A.D. WOO to 1900. Tellem is usedwith two meanings: a) for all pieces before c. 1450 and b) for all
period, whatever the
style shown by Plates 2.5 and 2.6. Confusion results from a label that defines by time
date, of the
style

and genre on

the one

hand and by genre only on

the other.

Only dated

pieces are best called Tellem

It may still be possible to date some works in wood

one day by the tree-ring method.
The best known of such techniques is carbon 14 dating. It measures the
decay of the carbon 14 isotope in organic matter, a decay that occurs at a steady
rate. Thus, objects that date from 250 to over 50 000 years back can be dated,
provided they were made of organic materials. Apart from test objects from
Pharaonic Egypt, only a sample of wooden statues from the Bandiagara caves
in Mali (Tellem) have been dated so far (by T. Northern). But the method is
very widespread in archaeology and provides the chronology for African sites
down to c. 1700, and at the same time for art objects associated with them as at
Igbo Ukwu. Most iron objects older than c. 1700 can also be dated by similar
means (Van der Merwe 1969), but the actual technique used to measure

well as on old iron objects.

amounts of carbon available is still experimental.

Thermoluminescence dating of ceramics relies on the

fact that

energy

accumulates in quartz (sand) grains in the clay as a result of bombardment of

particles produced by the radioactive decay of isotopes of uranium, thorium
and potassium which are also naturally present in clay. This energy is released
in the form of light when the quartz grains in a sample taken from the object are
heated, and can be measured in the laboratory. The amount of light emitted is
proportional to the time which has elapsed since the object was fired. The
method has become quite common during the last few years and has been used
to date not only terra cotta from the Middle Niger but also copper alloy
artifacts
by measuring their clay cores (Stoneham 1980; Fleming

1977:193-200).
A technique now being developed would also allow us to date paint. Rock
paintings can now be dated by the rate of rearrangement of aminoacid
molecules in the binding medium of the paint. But so far the precision of the
results is not great enough to be very useful since the standard deviation still
38

IDENTIFICATION
runs to several hundreds of years (Denninger 1971; Iskander 1980: fn. 46;
Phillipson 1977:268-9).
AH of the extant techniques only yield approximations expressed in
numbers and standard deviations. For radiocarbon dates these drawbacks can
be overcome by relying on block dating, i.e. obtaining a number of dates (at
least three) for closely related material. Particular deviations of results,

stemming from

variations in the

raw materials (pattern of growth in wood,

and perhaps the locale (corrective

specific properties of shell, bone, etc.)

curves proposed for different tropical areas) are often not very significant, but
can and often are taken into account to reach greater precision.
So far, then, only items in stone, textiles and non-ferrous metals cannot
be dated. It is not impossible, however, that techniques for dating
will be found even for them and certainly museum curators should search out
possibilities for developing and testing such means. As so much African
sculpture, however, is carved out of wood and is presumably younger than
1700, other dating techniques should be developed. Already Olbrechts cited
dendrochronology, or the tree-ring count method, which yields precise dates.
The method is well estabhshed and still expanding for woods in temperate
climates. With regard to the tropics, however, quite a controversy rages. Most
botanists argue that where tree-rings occur, even in tropical conditions, the
method should be applicable (Doutrelepont, personal communication 1980).
But banks of comparative natural materials must be set up and quick results

cannot be expected, especially since local climatic effects require many control
banks rather than just a few for the continent. Opponents claim that it cannot
be done, because uncertainties would be too great, because the results would
not be worth the effort, and - among museum curators - because works of art
should not be mutilated, even by taking fractions out of their core. The issue is
far from settled. Given the crucial contribution that tree-ring counts could
bring to chronology, certainly in the drier parts of Africa, and even for some
species in the equatorial forest, one hopes that efforts will be increased.
So far art historians of sub-Saharan Africa have remained defeatist with
regard to dating, especially when it comes to wood. They argue that tropical
for long and that every piece that entered collections
after 1900 is practially contemporary with the generation of the sculptor. In the
tropics wood decays fast, even if the objects are seemingly cared for. In practice

woods cannot survive

is very difficult to accept and certainly remains unproven. Some objects

were and are kept with great care and might be much older than others. But
there is no way of knowing at all unless and until physical means for dating
have been perfected to cope with this problem.
Approximate methods of dating by the object in the absence of any
outside help consist of relative ranking by style, which will be discussed later,
and in dating by context. An example will make the principle clear. A Benin
royal head cannot antedate kingship. The presumed royal crown of Loango
cannot antedate kingdoms in that area. But one cannot claim the contrary, that
is, that this crown dates to the foundation of the kingdom in the fifteenth
century or earlier (Volavka 198 lb: 51). Obviously the crown can have been
fashioned later; it only dates to sometime after the kingdom was founded. Such
reasoning, which depends on use, is always risky, however, because use can
change. Thus, in Coptic Egypt, existing erotes figures were unchanged but

this

39

ART HISTORY IN AFRICA

came to mean

now

'angels', while

heathen belly dancers continued to be carved but

carried crosses over their heads.

CATALOGUES
has been the practice to date pieces by comparison just as they have been
assigned places of origin by comparison. In the absence of any other data, this
at least is a first step to bringing order into a mass of art objects. To do this well,
however, a systematic reference catalogue listing all known objects and all
iconographic representations for different periods of time would be necessary.
It

Thus, one could attempt to reproduce all known objects from sub-Saharan
Africa in Europe, acquired before 1700, with details about origin in one or
more volumes. These would include most of the Afro-Portuguese ivories, but
also other ivories, works in wood, textiles, etc. It would, on the other hand,
make clear which works were actually dated and which are only dated by
association, and how works were dated. Moreover, such a volume could refer
to the relevant texts in documents of the period. Standard reference volumes

up to perhaps 1870 or 1900 could be built up. The

would not only help in placing and dating undocumented objects, but
give us a firm basis for a first general framework in art history itself. If this were
to be done, one would discover that there are fewer eighteenth-century objects
in Europe or America than seventeenth-century objects, that some objects
have been crucial for dating many others, even though their own dates are
rather vague, but again we could also expect some pleasant surprises.
There is no way that this task can be eschewed if the field is to develop.
Perhaps it will not occur in quite this fashion, but rather piecemeal by type of
object and medium through series of monographic studies. But an inventory of
what was extant in Europe before c. 1880 will have to be made. If, in addition,
the tree-ring controversy could be solved and physical dating became more
for succeeding periods,
result

routine, the total picture could be transformed. To achieve this end, however,
the very first requirement is to revive an awareness of the crucial role of
chronology and to have more faith in the potential results of dating techniques.

There are encouraging signs that such

40

a revival is

underway.

CHAPTER THREE

SOCIETY, THE MOTHER OF

ART
Works of art

are not symbols only, but objects in the true sense, necessary to the

life

of

social groups

(Francastel 1951:8-9)

OBJECT AND SOCIETY

and not only as symbols but as objects in the
of social groups. Hence activities of daily life
such as farming, fishing, hunting, cooking, even walking, use objects and not
all of these are tools in the narrow sense. Charms for insuring success in any
activity, for example, are not tools, but they are utilitarian. All passages from
one condition of life to another use objects to mark these occasions and the

All social activity uses objects,

true sense necessary to the

resulting

life

Birth, initiation, marriage, and funerals are often

special objects used in special action. Crises such as war,

statuses.

surrounded by

detection of witchcraft or curing use their

own

assemblies, spokesmen for communities and rulers

objects.

all

have

Courts of law,

theirs.

Games and

entertainment had their own objects. Social status thrives on distinctions of

dress, finery, food, furniture, transportation, and housing. Objects not only
distinguished between social categories but also between social roles. The
headcover of a Muslim judge is not that of the reader of the Quran in the
mosque. The emblems of the woodcarver are not those of the smith. Objects
were made for use, had social meaning and cultural value. They were almost

never made, however, only to be an expression of art for art's sake. As tools and
as symbols, objects reflected every facet of the

An inventory of all the objects

community.

is a description in filigree
of that society itself. Inventories are the material precipitation of social life in
all its complexity. This is why archaeologists can attempt to reconstruct
long-dead societies and cultures from objects. Art historians are in a similar
position if they have access to information relating to use, institutional context,
persons and groups involved, and workshops. But their primary task is truly to
understand a work of art. An objet trouve without context cannot lead them
beyond supposed use. To understand fully a work of art, the social

used by a society

its creation, use and function must be known.

which art is formed. Therefore, aside from aesthetic
general visual and tactile nature, the social context in which it is

circumstances surrounding
Society

is

the crucible in

elements of a

is crucial to art history (Brain 1980). In this chapter I will first deal with
the use, institutional links and functions of works of art, and then turn to the

created

social matrix of its creation: patronage,

workshops and product.

41

ART HISTORY IN AFRICA

USE
Most objects are used for what they seem to be used for; they are utiUtarian.
Door locks, heddles, pipes require little by way of explanation as to use;
however, one may wonder about their shapes and especially about the
differences between ordinary specimens and those that have been made works
of art by carving or by decoration. Thus, the lid of a Kuba cooking pot,
begrimed with soot, is clearly used to cover the pot, but why would it be plaited
in intricate decorative patterns that are quickly obscured by the soot? It took
to three weeks' work to fashion such a lid, and yet soon the charming
decoration became invisible. Why then was it decorated? The lid was an
indication of status, and reveals that wealthy people could afford the
expenditure for labour required to make such a lid.
The actual use of many objects is not self evident and errors of assignment
have been made. Thus a copper object, labelled 'fishing basket' in a great
museum (Plate 3.1), was later believed to be the crown of the monarchs of

up

Loango (Volavka 1981b)! Other scholars viewing it with the objects associated
it believe that the object was a shrine. Among the hypogee rock-hewn

with

structures at Lalibela in Ethiopia, two that were believed to be churches like

1968:104-5;
(Gerster
palaces
were
they
not;
were
others,
the

the dome shaped object which has a large hole near the apex. Fifteen
NearKabinda, Angola. Musee deL'homme. Diameter 36 Scm; height 20 cm (double the
The metal
size of a chiefs cap). Before 1933. The museum, after Tastevin, calls u a shnnefor Lusunzi, spirit of the fish.
as
objects were strewn around the dome and represented the fish Volavka (1981b) sees the objects as regalia and the dome
the
the crown of the old kingdom of Kongo. She dates the dome to the 13th or 14lhc. when that kingdom arose. The size of
dome makes it too big for wear. Crowns do not necessarily date from the foundation of kingdoms

Plate 3.1

Crown

or shrine?

copper and iron objects with

Copper thread for

it.

42

Plate 3.2 Figure. Soapsione. Found in soil and re-med. Original use
unknown. Sherhro Peninsula, Sierra Leone. Hntish Museum. Height ?5mm.
Acquired 1904. Relieved to he 15th or 16th c. Such figures are labelled nomoli

43

ART HISTORY

IN

AFRICA

Leroy 1973:145-6).

On

postcards a circular-looking building around an open

and adjacent parts of Libya is labelled

oval in the interior of southern Tunisia

'Berber Palace', whereas such structures, in fact, are granaries.

Moreover, questions of use, when art is involved, can also have
complexities. Thus, theminbar in north African mosques is a staircase, leading
to a platform. Obviously its use is not that of an ordinary staircase and no one
ever ascends the highest rungs. The imam addresses his congregation on

Fridays from one of the lower steps. Consideration of use related to structure
such as: why so many steps? Why the shaping of a platform on
top that is never used? Why an entrance gate to some, not to others? Why is the
minbar on wheels in northwest Africa and not elsewhere (Schacht 1957)? Often
raises questions

be couched in language placing so-called functional features in

opposition to non-essential features. It should be stressed that 'functional' here
refers not to function (the overall impact of the object on the community that
keeps it or/and its goal), but to use. We can only know what is 'functional' in an
object if we know how it was actually used. That is the first step for any
a discussion will

understanding of any object, work of art or not.

Although such an observation may seem banal, it warrants attention
because, in many cases, the precise use of an object cannot be determined for
lack of information. This obviously is true for many objects found out of
context such as the stone figurines in Sierra Leone or Guinee. Local finders put
them into fields to raise an abundant harvest and sometimes whipped them
when the rains were late (Paulme 1981). But how had the makers used them
they put in shrines and talked to? We do not
not a simple concept. Objects such as decorated
flywhisks are emblems of authority. They are used by the holder to underline
his status when addressing a public or presiding over a court. If in so doing, he
several centuries earlier?

know. Moreover, use

Were

is

swats a fly, that is incidental. The main use, like the main use of drums of office
or of thrones, rather than stools, expresses social relationships. Unlike tools or
utensils, such objects are not strictly utilitarian. Such a non-utilitarian use has
often been called function, but as

another sense. Function

is

we

shall see that

more than observable

term

is

better applied in

use.

PATRONAGE*
Most works of art in Africa were commissioned for individuals or for collective
groups. Artists on occasion also worked for themselves to produce art works as
utilitarian objects, to advertise their mastery, or in

even rarer cases,

just for

fun

some person or event near and dear to them. But

usually there were patrons and commissions.
The majority of patrons were also the consumers or represented a

or to

commemorate

privately

consumers. Individuals often represented collectivities as when

commissions came from bishops ,GaJi'5, abbots, sultans, kings, village leaders.
The works commissioned were most often destined for their collectivities,
although they also served more personal needs of the commissioner, for
example, when a palace had to be built. Individual patrons quite often
commissioned works for the collectivity as well. This was the case with the
collectivity of

See particularly Ben

44

Amos

1980b.

SOCIETY,

THE MOTHER OF ART

who built a commemorative mosque or the

man who had broken taboos in southern Cameroun who

successful merchant or general

wealthy

commissioned the so initiation ceremonies for the young and the sculptures
that were part of it (Laburthe-Tolra 1977: vol. 2 1359-1554). Sometimes the
patron did not specify the details of the commission, leaving the decisions to
someone else. Thus, in eastern Liberia, the^o religious master decided who
would be entitled to a new mask and what type of mask it should be (Gerbrands
1957:86).

In practice the community of patrons and users expected works to have

prescribed iconographies executed in a certain style. Works that were not up to
expectation might be refused or, for assurance, patrons might supervise the
execution as in the case of buildings. The conception of a work of art thus
always involved the community as much as the artist in the same way that an
interview is the joint product of interviewer and interviewee. Even when the

customer was unknown, the

for sale to

would

artist

unknown patrons, he

was still influenced. Even if he made objects

had to know what they liked (that is, what

still

sell).

In this sense, then, works of art are truly collective creations. The
magnificent frescoes at Faras, for instance, were not conceived exclusively by
the artists. Each fresco, each theme, was predetermined for each portion of the

common

whole eastern
were proper for
crucifixions, for the positioning of archangels Michael and Raphael, for
various scenes from the liturgical calendar. The local artists at Faras followed
wall of the cathedral church, according to rules

Christian Church.

The patron/bishop knew which

to the

locations

convention for centuries in these matters of location, as they did in matters of

themes and composition. The artist's freedom in such matters was limited.
On the other hand, while the collective character of works of art is true
enough, it has often been misinterpreted by partisans of 'tribal art'. They
stressed the dead weight of tradition and claimed that artists were but artisans
lacking all creativity. We know that was not true for Pharaonic Egypt or for
Eastern Christianity, or for the architects of mosques in the classical style,
although traditions certainly were as stringent, and probably more so, than in
regional traditions. There were more dynamics in the evolution of taste in
collectivities than the partisans of tribal art were willing to acknowledge,

probably because they underestimated the complexity of the relationships

practical and aesthetic - between artist and public. In Africa south of the
pubhc commissioned and the artist
public
by introducing innovations in the
his
of
could attempt to revise the taste
Sahara, as in

work

that

It

many

instances elsewhere, the

was commissioned.

was

single style.

rarely true that a given ethnic

Often different

group was completely wedded

styles coexisted in the repertoire,

to a

being applied to

when in southern Gabon naturalistic masks were made for

same public by the same workshops which crafted highly expressive
ancestral figures. The extremes of angular cubism and smooth affected
naturalism were turned out by the workshops of eastern Liberia according to
the demand. It is incorrect in a case like this to say that they turned out Guere
(Wee) masks or Dan (Gio) masks as if Wee patrons asked for one type and Dan

different objects as

the

The same clientele could ask for one or for the other. In
mosque could sport a Turkish minaret next to a main

patrons for the other.

northern Africa, a

45

ART HISTORY

IN AFRICA

The expectations of patrons, users and artists did not

and innovation. Within given ethnic groups, different
workshops could have different ideas about what form was proper for what
object and in some cases form did not matter much at all, provided a very
specific meaning was expressed. Initiation baskets of the Bwami association
(Lega, Eastern Zaire) contained well-carved ivory and wooden artifacts of
great diversity along with buttons, celluloid dolls and natural objects. Whether
the thing was an apt metaphor expressing the intended proverb or aphorism
mattered here; not its style (Biebuyck 1973:142-57).
And yet specific art forms were related to specific communities, often
dominant in a society. Thus, there could be a court art in kingdoms such as
Benin, where specific objects with clear stylistic characteristics were reserved
for the king or the court and could become items involved in competition for
power; and there could be court art as there was among the Kuba, where the
most appreciated and imitated works gravitated towards the courts which
remained the pace setter of fashion. In such cases, folk art, in the sense of
provincial imitation, existed. For example, a small nineteenth-century
Moroccan rural fort (Ksar) sponed an imitation of the celebrated gates at
Meknes, while others copied the famous mihrahs of the great mosques at Fes,
no doubt as an expression of local pride. Folk art in another sense existed
where skilled artists could not be afforded by poorer folk, who imitated the
better crafted works done for their superiors. Kuba dolls are an outstanding
example of this; some are fine carvings, some crude toys.
building in classical style.

rule out flexibility

Artistically, the taste of social classes could differ very much within a
given society. In Kongo in the sixteenth and seventeenth centuries, the upper
class was Christian and adopted Christian imagery and indigenized variants of
Renaissance styles, while most of the population remained faithful to its own
art objects and their styles. The East African coast with its division between
ruling strata and their Islamized tastes as against the other layers and their
'African' tastes is another case in point (de Vere Allen 1974). The very
is based on social stratification. At first Copts were
Egyptian Christians as opposed to the higher classes who were either devoted

definition of Coptic art

to Hellenistic art or to early Christian Imperial fashions. After the

Muslim

conquest of Egypt, Coptic art was the Christian art of the country, still
practised by the lower classes and by isolated monasteries for centuries (Wessel
1965:53-80; Du Bourguet 1967:3-15). The more research has progressed the
more such differences between social strata have become apparent. Further
examples could not only be adduced from the oikoumene, but from places as
diverse as the upper Middle Niger, the Akan area of lower Ghana and the
peoples of central Angola. Tastes may not have varied as much there, but
specific representations and the quality of the work attests to stratified
societies.

Clearly ethnicity has been overestimated in the past as a force shaping

forms of art and class has been underestimated. But it would also be wrong to
see class as a dead hand stifling innovation in the same way as the partisans of
tribal art viewed the relationship between ethnic group and artistic expression.
The relationship rather revolved around the use of art objects linked to
institutions which in turn were themselves linked to class. Thus, in Kongo,

East Africa, and

46

among

the Copts, religious institutions requisitioned the art

SOCIETY,

THE MOTHER OF ART

the styles. At the Benin or Kuba courts, the taste of the patricians
prevailed and expressed itself usually in the formula 'rich is better' and 'more is
better' promoting rare media such as ivory or even gold and more lavish

made and set

decoration.

CONTEXT

SPECIFIC SOCIAL

and art is the primary one. The

on form which institutions
exercised. Given a similar basic shape of drum, three drums are compared: a
village drum of the Kuba, a royal drum of the Kuba, and a village drum of their
western neighbours, the Lele (see Plates 3.3 and 3.4). The difference in
execution is striking. The Kuba drum, owned and used by the village, is
decorated only with a modest band of decoration. The Lele drum usually
shows decorative carving all over in a very fine pattern and exhibits a human
face on its side; this rich drum is also only village owned and village used. But
Lele villages, unlike the Kuba villages, were sovereign units. They were often
larger and always more proud and their drums show it. The Kuba royal drum
has deep incised decorative patterns all over and inlays of copper, beads and
cowries. It was much richer than the Lele drum and reflected the institution of
kingship even though a dynastic drum such as this one was the emblem of one

The connection between

following example

social institutions

illustrates

the

influence

king only, not the drum of kingship itself.

On examination, both the ownership of works of art and the conditions of
their use show which objects belong to which institutions and are used on
which occasions. In the case of architecture, it is particularly easy to

communal and private dwellings, to discern what are

temples for spirit worship, palaces, men's clubs,
shrines,
churches,
mosques,
and so on. In the case of sculpture and painting, textiles, or body arts, the
difference between personal belongings, with or without a mark of social class,
and collectively owned or used art, as in masks for instance, also becomes
evident. It is possible in this way to express a meaningful link between an art

distinguish between

object and

These

work of

its

social niche.

form of a work of art as well. Usually one type of

defined usage within an institution. This tyj)e of

links extend to the

art is

found

for a

an ancestral figure, that type of statue is a charm for healing, this type
of mask is used at such and such a stage in initiations, that type at another
stage. A one-to-one correspondence between form and use in a given
institution, nevertheless, did not always obtain. Masks could be used in other
contexts than the primary one for which they had been fashioned. This has
been reported for eastern Liberia, where masks of similar appearance could be
used together in different ways, within the same general institution, according

statue

is

and standing of their owners (Fischer 1978). In other ways

masks or other objects could be used outside their primary context among the
Kuba where the primary context of some masks was in mimed
dances at the capital, ahhough they also appeared on the top of initiation
walls. Nevertheless, there rarely exists a problem in identifying the
primary context and the relationships between works of art used in a given
community or between works of art and other objects there.
Because the tie between institution and work of art can easily be

to the prestige

47

ART HISTORY

IN AFRICA

Plate 3.3 Roval drums, pel ambish. VCoocl. copper,

musees nalumaux du Zaire. Heiuhl

1 1

anmet,

heads. Nshenf;,

(km; diameter Mhm. 19th c.

Kuha kingdom. Noiv

he main pattern on each

drum

al Inslitut des

tdenlijtes the

kmfijor

WON made (one per reii;n). Rams' fwuds under the knohs. The richest decoratums arc the latest drums. IHth c.
drums ojlhe same shape and aeneral composition have no heads nor cowries andimlv a sini;le copper plaque incruslatwn.

iL'hom

It

Kinf;s associated hy oral tradition

48

SOCIETY,

Plate 3.4 Village drum. Wood.

I.ele village, Zaire.

Komnkhjk Museum

28cm. Date of acquisition: 1924. Probably 20th c. Such drums

head (a few exceptions are known) and ornament in very low reliej

ividih

THE MOTHER OF ART

voor Midden Afnka, lervuren. Height 92cm;

have a similar composition with the carved

all

49

ART HISTORY

IN

AFRICA

only by an examination of the ownership and control over use of

the work of art, a nomenclature based on the institution to which a work of art
is linked, rather than on a simple ethnic name, is much to be preferred. Thus,
ancestral figurines of northeastern Gabon should not be labelled 'Kota'. They
are figurines to be put on top of bundles or baskets containing the relics of
ancestors; a short label would be: 'Ancestral relic figures'. The geographical
established,

if

location should be expressed by village or workshop, where known, or else in

general terms as 'northeastern Gabon'. In this way the nomenclature would
avoid the false problems that arise out of tying ethnicity to certain objects and

certain styles.

ARTISTS

AND WORKSHOPS

which works of art were produced are the workshops. Even

an artist working alone constitutes a workshop. It is there that youngsters learn

The institutions

in

the craft as apprentices,

it is

production take place, and

disseminated. Artists were

there that the crucial portions of the process of

it is from there that the finished product is
artisans and to the public their skill or

was what made them different from other people. As in

men and women put their hands to many different
tasks. South of the Sahara, wood carving or weaving were known to most men,
while most women could fashion pottery for instance. Hence one hears of men
crafting dolls for their children or medicine men carving their own charm
specialization

preindustrial societies, most

statues.

But

skilled

work remained he

attribute of the specialist, the artist,

because he had more practice.

Almost everywhere the division of labour between the sexes left
metalworking, building, carving to the men, while pottery was often for
women and weaving was man's work south of the Sahara, but woman's work
for certain textiles in the north. Often women had something to do with the
decoration of woven stuffs. In rare instances women were woodcarvers as well
as men. The only reported situation, so far, where they could be equal is in
eastern Zaire, just east of the Upper Zaire among the Hemba speakers (Neyt
1977). But women as makers of ceramics, mud sculpture, sculptures with
paste, decorators on walls or on textiles were not rare. Still there is no denying
that those arts which were felt to be important by men were reserved for males
almost everywhere (Gardi 1969).

Like

all

specialists artisans started out as apprentices.

Often they were

of blood, but equally often they were

related to the master of the shop by
masters on the strength of promise.
unrelated
by
accepted
been
not and had
Beginning with the most menial and crudest of chores, apprentices gradually
ties

learned the trade by imitation and they advanced in skill a step at a time.
Depending on the media and techniques of the art, this took a shorter or longer
time. Thus one became a painter in Ethiopia after a longer period of

apprenticeship than an illuminator.

Workshops were large or small and this depended in part on the
techniques involved. A building site contained many people, while a carver
might work alone or just with one or two apprentices. In some cases artists
were grouped in guilds, the most famous instance being that of the ivory and
the brass guilds at the court of Benin. They could only work for others with the

50

SOCIETY,
king's approval

and were kept under

THE MOTHER OF ART

his watchful eye in the capital

(Ben

Amos

1980a).
social status of artists varied from society to society. Where manual
was
despised most, visual artists were despised as well. Thus in western
labour
Sudan, Ethiopia and northern Africa, metalworkers, carvers and workers in
leather were casted; they could not marry people outside their caste and their
calling was strictly hereditary. This did not necessarily mean that individuals
were not respected for their skills and their creation, but in the social order
their status was low. In Fatimid Egypt, Copts were weavers, woodworkers,
masons, potters and architects for minor projects. They came from the lower
classes and as such were not prized. Those who had remained Christian but
had risen in society dissociated themselves from the artists. Other Copts were
monks or attached to monasteries like many artists in Nubia and Ethiopia.
Their status as artists was not different from the status monks enjoyed in

The

general.

In portions of Africa some forms of manual labour were highly prized.

Smiths enjoyed the highest esteem in west Central Africa so that kings
pretended to be descendants of smiths or at least learned how to work iron (de
Heusch 1956; de Maret 1980a). Kuba kings all claimed to be carvers and the
visual artists in that kingdom were held in high repute, although they did not

enjoy special privileges. Chiefs among the Bangwa, a portion of the grasslands
of Cameroon, thought so highly of carving that they maintained the fiction that
they were the artists, but thought so lowly of carvers that most of these were
slaves attached to their household (Brain 1980:135-6). The workshop here
produced for the king who then gave away or sold the carvings as his own
creation! From the above it appears, then, that no special status or role was set
aside for the artist as artist, but that his place in each society was that of the
appropriate category of labourers.
Social conventions also moulded the activities of workshops. Some were
more specialized than others. Thus in West Africa's Sahel, smiths were also

woodcarvers and their wives were potters. The same workshop produced all
some smiths were potters but not carvers and
their workshops would be different. Most of the smiths' wives, however,
would be potters, so that among pottery shops some were thus linked to the
smithy and others were not.
The production programme of a workshop could be very varied or very

these products. In Central Africa

restricted.

where specialization in general was not

workshop made masks, carved stools for boys'
carved spoons, huge display spoons for the wives of leaders,

Thus,

in eastern Liberia,

greatly developed, a carver's

initiations,

dancing staffs, staffs of office, birds as gable ornaments, little animal figures to
be given to the village headmen as a token of respect, wooden statuettes,
neck-rests, playing boards for warn, pipes, house-posts and a host of
implements (Himmelheber 1960:164-81). In the Kuba capital on the other
hand the specialization of labour in general was very pronounced by the 1890s.
It was natural to find workshops there that did nothing else but carve the heads
of pipes and others merely the stems! The production programme was in part
the result of technical processes and in part the reflection of the general
division of labour.
It is

not surprising that

when

it

came

to the disposal of the products, the

51

ART HISTORY

IN AFRICA

workshops acted like any other production centre. In kingdoms they were
taxed on labour or by being given commissions or even on products made for
sale. Products were made for sale in many societies, especially when they were
unique in some respect, so the art historian must take into account the
possibility that some shapes and styles spread simply by trade. It is well known
that much of West Africa's decorative arts have been influenced by Islamic
models from the north and it is also well attested that many of the examples that
were copied moved through the trade of textiles and metal vessels in these
areas.

value of the products should never be forgotten. After

all, workshops had costs of production and these had to be met. The value or
'price' of the output was related to them and the greater or lesser
commercialization of the societies in which this art appeared established how

The commercial

costs could be calculated

and how they could be compensated. In general,

smiths were richer than the average person because their wares were in high
demand, but carvers were not, especially in societies where leaders were few

group was small. For some products the specialization was

so successful that production ran to an enormous output as with the celebrated
Tunisian glazed zallij tiles, or the no-less famous Coptic textiles from Egypt. In
other cases production had to be small. There were never more than perhaps

and the

size of the

150 or so frescoes executed in Christian Nubian churches of the Faras diocese.

The scale of the community was small and demand was slow. Whether a
permanent workshop could even function there remains unknown, but can be
questioned. In turn, raised output could lead to advances in techniques or
simply to a faster change in style, whereas the contrary could occur with a very
low output. It is therefore important to know of the high outputs of smiths in
Ghana who over several centuries turned out millions of goldweights.
Enforced idleness never occurred so their skills remained high and the
incentive to innovate was strong, if only to escape boredom - at least if the
public accepted innovations as they did there.
The economics of production are important in any consideration of art.
The volume of output, the size of workshop, the size of the area affected by the
product, the possible speed of innovation in form, the relative number of

and even the relative

there would be few
skilled specialists in a rare and costly medium and few commissions. The
patrons, affluent leaders, could influence shape more perhaps than in cheap
work for large markets. And certainly such rare works set the dominant taste
and affected general stylistic evolution as their example trickled down.

workshops, the

relative

frequency of skilled

influence of the taste of patrons,

all

were related

artists

to

it.

Thus

FUNCTION
Having sketched out the relationships between works of art and society in
general, we can now turn to a consideration of what social scientists call
function. Unlike most art historians who mean by this 'use', they apply the
term to the effects of use on social relationships. They see art objects as the
crystallization of social relationships and a tool in social communication. Art
can express power, status, wealth, challenge. Function therefore differs from
use or goal as

52

it is

the expression of an effect.

Thus funerary

pole carving in

SOCIETY,

THE MOTHER OF ART

Mijikenda country behind the Kenyan coast was used to mark a grave and its
was to commemorate ancestors (Brown 1980). But its function was
polyvalent. These carvings helped to assert the authority of elders over youths,
to enforce customary norms left by the ancestors over innovations and the like.
Through its link with ancestor worship it affected all social relationships in that

goal

society.

The

fact that expressions of visual art are objects, concrete things, is

For unlike mere mental propositions, such concrete

especially important here.

visual concepts as are expressed in art acquired great reality, as

we shall discuss

and hence their effects were enhanced. This remains true even for
ephemeral products such as masks or costumes used as props in performances
later,

of dances or festivals.
But the concept of function is treacherous. It implies a double cultural
interpretation: that of the community that uses the object and that of the
culture of social scientists who interpret the object in terms of institution and
social integration. Functions are always deduced, never observed. Because
meanings of art objects are often multiple, functions are multiple. Thus

mosques represent the community of worshippers, its dependence on God and

the rule of its law, but they often also unified a fraction of a community against
others who had their own mosques. They could express the importance of

when a benefactor built them. They expressed power in the case of all
Friday mosques, because that is where the prayer for the sovereign was
recounted on Fridays before the assembly of local government. Similarly, one
can say of masks in eastern Liberia that they expressed collective coercion, the
power of arbitration, the state of competition, the notions of healing and so on.
They were so polyvalent here that they in fact expressed all aspects of social
life, and their functional effect was diffuse but generalized.
Art historians should not accept the social scientist's function as a 'fact'
but as an interpretation. They should require convincing argumentation and
documentation and they should never forget that one function never excludes
another. In general, discussions about function often remain loose statements

wealth,

between art, society and

can make up their own
express such relationships.

turning around the nature of the most profound

culture. If the

documentation

minds and marshall

their

is

adequate

ties

art historians

own knowledge

to

DOCUMENTATION OF SOCIAL CONTEXT

it is, social context is not given by the object but only by descriptive
accounts that stem from local authors or foreign observers. Even in northern
Africa, where many local writings are extant, much still stems from foreigners,

Crucial as

only perhaps because foreigners were struck by practices, institutions and

works of art that were unremarkable to locals. By far the majority of
sources stemmed from men and most foreign accounts were written by
professional travellers, businessmen, military officers, civilian administrators,
missionaries, physicians, transitory persons such as sailors or Christian slaves
in North Africa, or even, from the 1870s onwards, anthropologists (Pechuel
Loesche 1907). Local sources were usually chroniclers although a number of
if

related

letters

and accounts can

The

also be

found (Djait 1980; Hrbek 1980).

usual historical critique should be applied to

all

these sources. First

53

ART HISTORY

IN AFRICA

the distinctions between primary and secondary source

is

highly important.

Ibn Battuta (1352-3) did travel to Mah whereas al Bakri (1067/8) relied on
merchants' reports. A notice in a museum, or even worse in a general printed
work about African art is derivative and its source should be sought.
An example is provided by a Kuba object representing a folded hand
(Plate 3.5). This was said to be the emblem of a society of warriors and to
represent the hand of a slain enemy in a number of books, all copying one
another. Indeed, W. Fagg who had visited Kuba country himself said this in
1958 even for a hand carved on the handle of a drum, quite a usual way of
dealing with the handles of a type of village drum (Elisofon and Fagg 1958:211,
n.270). I was told that this merely referred to the handle as a visual pun. None
of the books gives its source of information, but they obviously copied one
another. The first ethnographer of the Kuba, E. Torday, did not mention this
in his works, and during my fieldwork (1953-6) it became evident that no such
association ever existed and the motif did not represent the cut hand of an
enemy. Nor did Frobenius, who collected the earliest such object, give this
meaning. He merely called it a ritual object (Krieger 1969 vol. 3:66).
The earliest reference seems to stem from an administrator in the early
1920s and came from a file at the museum of Tervuren (Belgium). No one ever

in shape of hand and forearm. Wood, iron hook at back. Use unknown. Northern Kuba. Museum fiir
Volkerkunde, Berlm-Dahlem. Height 4S-5cm. 1906. Probably acquired by Frobenius at Bolombo on the Sankuru nver.

Plate 3.5 Emblem

For

54

his

method of acquisition

cf.

Frobenius 1907: photo 453

SOCIETY,

THE MOTHER OF ART

The mention of a 'secret society' by Fagg is not very

His stay was very short, this clearly was an incidental comment
and he may have been misled. The cut hand motif does go back to the
Independent Congo State and many, in this area, refer to a raid in 1899 by allies
questioned his

reliability.

reliable either.

of the State on a portion of the kingdom. The eye-witness description of the

raiders smoking cut hands over a fire stems from an observer who had been in
the country since 1892 and wrote much about the Kuba. The very fact that he
did not mention any such practice among the Kuba is more significant than
negative evidence usually is (Elisofon and Fagg 1958: no. 270, 211; Preys n.d.;
Shaloff 1970). The whole attribution is clearly a fabrication in one or more
stages.

But the prevalence for sensational labels such as 'sacrifical stool',

and others, indicates that such inventions have not been

'executioner's knife'

unusual

at all!

When using documentation it is important not only to establish whether it

stemmed from an eyewitness, but equally to assess the credence to
With regard to foreigners, obvious points to check
whether they could understand what they saw. Hence the length of stay,

ultimately

give to such eyewitnesses.

are

the degree of fluency in the local language, the special interests and abilities of
the witness, the nature and the extent of his relations with the local population,
are important. The military were better on arms, the trader on trade, the
missionary on religion, but the latter was wont to misinterpret matters of
religion and certain details could be hidden from him. The military were often
waging war against the people they described, whilst the traders were often - in
the early days of colonial rule - obtaining goods by force. Anthropologists are
not necessarily better sources than others. Their theoretical concerns directed
both the selection of their observations and even more the selection of their
all

data for publication, as well as their interpretation of these data.

Internal sources include not only written material, to be examined by the

usual rules of evidence, but also oral texts. These can be precious provided
they do not refer to a long-past antiquity. Often they refer to the use of objects
and yield a fair amount of background about them, whether the narrative in
which they are embedded is itself supposed to be truth or fiction. A great deal

remains to be learned from such sources, not only narratives, but also proverbs

and songs.

Not nearly enough attention has been paid by specialists in the arts south
of the Sahara to such questions of documentation. Often even the elementary
levels of critique are not attained, and yet historical critique is crucial to
establish a sound body of data. To neglect it on the grounds that data from the
field are never forged is childish. Some totally faked relations exist and in many
cases it is not at all clear exactly how the data set forth were obtained nor how
reliable the observations about art actually were. To neglect to apply the rules
is only a minor by-product of an art
myopic. Given the primordial importance of context to
any understanding of any work of art, questions about the reliability of the
information should rank near the top of any list of concerns to the art historian.

of evidence on the grounds that this

historian's

main task

is

55

CHAPTER FOUR

MEDIA AND TECHNIQUES

make) in the literal sense of the Greek word. They
of
art should never forget it. An art work is the
are 'makers', and lovers
transformation of a medium into an artistic form, and each medium has
inherent physical properties which, in accordance with the techniques used by
the artist, affect the descriptive and aesthetic features of the work of art. They
do not merely allow it, or influence it, but they are form inchoate before the
work is done, form realized afterwards. Their qualities, shapes and textures
in another. A
differ, so that a work in one medium is unlike the same work

Artists are poets (poiein: to

grain,
bangle in ivory, decorated or not, differs from a bangle in copper, by the
ot the
shape
the
hint
at
the
by
more
even
but
touch,
the
colour,
the
the texture,
of metal
material: the natural hoUowness of the tusk as against the fluidity
to one
setting around another hoUowness. Visual images like all objects belong

medium, or to one set of mixed media. A raised tattoo is a form of decoration. It

cannot be imagined to have the same effect on any other material than the

human skin. Yet objects, even art objects, sometimes are carried out in another
medium than the usual one, just as an engraving differs from a drawing. The
Kuba imitate decorated calabashes in wood, but the result will never be
mistaken for the original. It becomes a new art object on its own, with its own
another is
values. This process of translating the form of one medium into
skeuomorph development.
Media themselves are often already transformations from

called

inert natural

not iron, the raffia

the potter
medium
palm frond is not the raffia textile, raw clay is not the
material
raw
between
intrude
already
handles. Often technological processes
we
chapter
this
In
artist.
the
by
executed
are
they
not
or
whether
and medium,

material into

man-made

materials.

The

iron ore block

is

cannot describe all technologies that transform raw materials into media,
because that would be tantamount to surveying the production of all material
the major
culture in the whole continent (Gardi 1969). We can only mention
for the
only
that
even
and
art,
of
works
become
media
which
by
techniques
of
most common resulting arts: architecture, sculpture and major forms
two-dimensional art.
medium,
Creative conception, coupled with skilful manipulation of the
of these
concordance
the
And
one.
fine
a
be
is
to
of
art
work
if
a
necessities
are
artisan.
indifferent
features marks the work of a superb artist in contrast to an
of the
south
But
cultures.
African
Virtuosity is highly appreciated in many
Sahara

transformation of a raw material into a very similar

example, in the case of metals and sometimes the making of

at least, the

medium,

as, for

the form out of the

observe
56

medium

is

not

all

skill. The makers must

The raw material can be

attributed to

rules relating to the 'real invisible' world.

MEDIA AND TECHNIQUES

guarded by spirits, as is the case for gold in the western Sudan; or it may be
either the abode or even the nature of a spirit, as when a tree is propitiated
before being cut; or the belief is not explicit but ritual purity is required during
the working of the material as when smelters must abstain from sex before they
reduce iron. In African cosmogonies raw materials, and the activities of
reducing them to media and then to fmd hidden form, have a place. No
'natural' action alone will produce transformations that are believed to be
meaningful. Making is not just prosaic, it involves invisible guidance and
inspiration. Especially striking new prime works are often attributed to the
revelation of form in dreams by supernatural beings to the artist.

ARCHITECTURE*
material for building was long-lasting, such as brick, stone, mud brick,
cement, or more ephemeral, such as mud, wood, fibres. It could be raw such as
undressed stone, but usually was made into a medium by dressing as for stone
or wood, baking (brick), puddling (mud cement), plaiting (fibres). Brick and
stone were the favourite materials for public buildings in northern Africa, mud
in West Africa, undressed stone, later regularly laid, and minimally dressed
stone in southeast Africa. Elsewhere wood, thatch, even fibres or leaves were
used. The impermanence of many architectural achievements south of the
Sahara has led to the popular belief that architecture did not exist there. That is
as erroneous as to claim that Japanese pavillions made out of wood and screens

The

are not architecture.

The techniques of architects, influenced by the purposes of the buildings,

were dictated by their goals. They were concerned with interior or enclosed
space and with the positioning of masses. In northern Africa architectural
concepts closely resemble European concepts. There technique coped with the
building of walls by taking standardized units such as brick, and by roofing
over the inner space by means of beams, arches, or domes. The basic technique
was additive. That means that the vision of form could be deployed in space,
unconstrained by the size of any block of material in which the form had to be
contained. Only one architecture in Africa was subtractive, i.e. hollowing the
building out of a mass given in advance. This is the case of the rockhewn,
monolithic churches of Ethiopia. The concept there is closer to sculpture than
to architecture (Gerster 1968).

The main problems encountered by architects stemmed from stresses

caused by the weight of the medium and by roofing resting on load-bearing
walls. The construction of openings such as doors and windows affected the
total strength of walls. Even blind walls were limited in height by such stresses.
As for roofing, both the span of units used to roof and the weight of the roofing
are classical problems. Wider spaces could be obtained by the construction of
internal supports (posts, columns, pillars, sections of interior wall) while larger
open spaces required structures such as arches (both in walls and between
walls), vaults and domes. Where the materials were lightweight and less
strong, the walls would not carry a heavy roof and interior space usually
remained quite constrained in width. Some remarkable achievements were
*

See Denver 1978; Margais 1954; Creswell 1952-9; Gerster 1968;

Badawy

1978.

57

ART HISTORY

IN AFRICA

|>v<'->.
58

MEDIA AND TECHNIQUES

nevertheless constructed as in the huge Mangbetu halls of the nineteenth
century with 12m wide barrel vaulted roofs or roofs sloping from c. 15m height
to very low sidewalls, while the front and back wall as well as interior posts

supported the armature. The length of such buildings was c. 30m and posed no
problem. Often architectural constraints were severe, because shifting
cultivation forced the people to move their settlements once a decade or more
solution consisted in building lightweight walls and roofs in
prefabricated form, so a dismantled public building could be re-erected

often.

The

elsewhere. In such circumstances, the architect's interest was not focused

much on interior space - most Hfe went on outside anyway - but on
relationships of buildings as masses to outside space, enclosed by walls or
streets. Thus, although the Kuba capital was not permanent, precise

measurements

for every street,

and

for the positioning of every public building

in or out of the palace, of every plaza, of every private

compound, were kept as

the architects, recreating similar effects wherever the capital

plan by
(Vansina 1976).
a

went

,C

Tm

m
3

^/

lozverpari doiunstream
1 Plan of the Kuba capital, Nshen);. 1953/6. Kuba ngk is our left. Upper part upstream,
and more prestigious. I Great square; 2 Square oj the crossroads; .? Square mbweengy 4 Square of the crown council; 5
markets; x public buildings; thick line:
Square of the roval dynasty; 6 Royal drum: the leopard in town'; 7 Yon drum;
Palace wall; thin lines: compound walls; dot: trees protecting the town. No details of palace grounds. Note, however, C:
daily council square; Y: private council square; Z; Harem square. In /95.? c. 2 000 inhabitants. In the 1890s between

Fig. 4.

'

000 and 10 000

Skill in architecture is a complex notion as building involves the

collaboration of many people: the architect, usually also the contractor and his
unskilled or skilled help. Skilled masons were quite important for stone, brick

mudbrick buildings. The effects of media on the form were obvious and not
always perhaps intentional. Intentional effects included ahlak in Mamluk
architecture, the alternation of stones of different hues, also a feature of early
Coptic churches in Nubia, or the patterning of bricks and stones as they were
laid. Less intentional effects, on the other hand, included the slope of

or

minarets, walls and buttresses in the

mud

architecture of the Sahel in

West
59

ART HISTORY

Plate A.lMinaret.

Top comers

60

IN AFRICA

Mud and wood.

in central

Saharan

Agadez, Xiger. Photo 1968. Reporied from the mid 19th

going back perhaps to 11th c. as a feature

style

c.

buirealage unknown.

MEDIA AND TECHNIQUES

Africa.

The visual effects imposed by the materials were remarkable there. Not

projecting
only do minarets and walls often look like hedgehogs with beams
horizontality
and
mass
of
impression
the
was
only
from them, not
overpowering, but details on big buildings or the whole of smaller buildings
than
such as holy tombs in southern Algeria, looked more like mud sculpture
volume
into
turned
mass
small,
quite
was
architecture. Where interior space
(see Fig. 7.1).

The hedgehog

effect resulted

from permanent scaffolding

for

walls needed. The sloping effect was a result of

coping with weight stresses and the horizontal effects stemmed from the
Low as the
inability to use such materials to build walls up to any great height.
buttressed.
massively
be
to
had
still
they
were,
Niger
the
on
mosques
of
walls

the continual repairs

mud

ADDITIVE SCULPTURE; METALWORKING

The

texture of a

medium

is

extremely important in sculpture, whether in the

space,
in relief, because just as architecture deals with the closure of
three-dimensional
a
it
is
architecture
sculpture is concerned with volume. Like

round or
art

but volume predominates, while mass in architecture

is

subsidiary to space.

Volume could be either built-up (additive technique) or subtracted (carving

will
proper). The differences in media, techniques and effects are such that we
or
ceramics,
metal,
usually
were
media
discuss them separately. Additive
powdered
of
made
Kuba
the
(among
pastes
as
such
materials
mud. Other
camwood) or even leather, such as in early nineteenth-century dolls from the
when the
Cape, were also additive. Additive media were quite pliable and weak
some
achieve
to
hardened
were
or
hardened
form was moulded and then
permanence. In

its

fluid or plastic state, the additive

medium

allowed the

making of almost any shape. The subsequent necessity was to achieve a

permanent stable state. Even mud sculpture must dry!
Metals have been used for jewellery and larger sculpture over most of the
African continent. The main metals were iron, copper and alloys of copper,
requires
gold and imported silver. Lead was sometimes used. Metallurgy
alloys.
of
preparation
the
complex expertise for the reduction of the ores and
southern
in
(iron
ago
millennia
two
from
Africa
Basic techniques were known in
and
Africa) to over five (copper in Northern Africa). The history of developing
varying technologies is still very imperfect, but the technologies existed and
produced different media. Thus iron is not just iron; steel, more or less
carbonized, has different properties of tensile strength and malleability. Cast
lack of
iron is a different medium altogether because of its brittleness and
to work.
elasticity, while wrought iron again differs by its ductility and ease
Already in the ninth century the smiths of Igbo Ukwu (Nigeria) knew that
copper was better for smithing and chasing while bronze was better for casting.
Leaded bronze is more ductile than copper, but copper is more easily
hammered, embossed, twisted and engraved (Shaw 1978:119).
The metal was worked either by heating and hammering into shape or by
heating it beyond the melting point and pouring it into a mould; that is, by
a hollow
casting. Open casts could be used, in which the mould was formed by
This
ground.
the
in
hole
stone, or by hollow ceramic, or even by a prepared
these
used
Zaire)
(lower
Kongo
use.
in
certainly was the oldest procedure
techniques, for example.

The

other main

method was lost-wax

casting. It

61

ART HISTORY IN AFRICA

Town.
Plate 4.3 Dolls. Leather stuffed with cotton. Glovelike stitch and material. Use: costume doll for sale. Cape
Museum fur Volkerkunde, Munich. Height 3Scm and 36 Scm. 1 81 5-1 830. European or European inspired work.
San,
Represents a San couple (man with cap). The tradition lasted at least until the late 1860s. Source for costume of
Khoi, Khosa and others portrayed, but usual name 'Zulu dolF unjustified

62

MEDIA AND TECHNIQUES

Plate 4.4 Vessel surmounted by leopard. Leaded bronze with low tin content. Isaiah-Igbo Ukwu. Length 20 Icm; width
94cm. Ninth century A.D.. The shell imitates that of a Tnton shell. Tnton shells are found on the coast lOOm away.
Note the antt-clockwise (sinistral) winding which is quite unusual. The very fine and dense decoration follows the

winding. The object

is

now

at the National

Museum, Lagos

(Nigeria)

involved the construction of a ceramic mould, the core, on which the icon to be
cast was modelled in wax. The core was then covered with a hollow clay mantle
and fixed to it to keep its relative position. The completed structure was then

heated up till the wax ran out, leaving an empty space in which the molten
metal was then poured. As wax is very pliable, the casting could achieve almost
incredible fineness and detail; but the process was complex. For instance,
special care had to be taken that the gases of a molten metal would not crack the
mould, but could escape through the outer wall, whose porosity was usually
increased for this purpose by incorporating organic material or charcoal into
the clay stuff that made up the wall. These tiny materials burned out and

opened the way

for the gases.

Ducts had

to

be provided for the wax and excess

and so that the stays that hnked

hamper
the flow of metal. Garrard
not
should
the core to the outer mande
(1980:116-21; Wilhams 1974:179-213) gives perhaps the most impressive
description of this incredibly difficult technique requiring perfect timing,
perfect control of temperature and an impressive deftness of hand. The
technique had been known in Egypt since Pharaonic times and spread to West

metal to leave the mould

at less visible points

Africa before the ninth century A.D.

Hammering and heating techniques also developed with diversified

operations such as annealing, hammering, punching, chasing, embossing,
filigreeing, and using inlays. Most parts of the continent show ample evidence
of a range of techniques for handling metals. Thus art objects are sources for
technological history and sometimes can be 'placed' by the techniques used. It
is

worthy of note that techniques of welding and riveting were not very

much
63

ART HISTORY IN AFRICA

became a permanent limitation of metal sculpture. The
Dahomey, where such techniques were used, are rare.
They can be contrasted to the seated statue of Tada (in the Ife style) which was

developed so that

size

giant iron statues of

cast in

one piece, an amazing

feat.

ADDITIVE SCULPTURE: CLAYS AND RELATED MEDIA*

Ceramic techniques had to be known before metal could be worked as tuyeres
the
for smelting and moulds for casting were usually made from clay. Basically
material.
the
homogenizing
grit
and
technique involved purifying clay, adding
a kiln. In most of
It was then shaped, dried and fired, either in the open or in
the
Africa, barring the northeast and lower Zaire after the sixteenth century,
or a
potter's wheel was not used. Bases were moulded on an older ceramic
natural base such as the bottom of a calabash. Walls were built up by coiling
That is the
rolls of clay and smoothing the walls to the profile wanted.
colombine method. Technical intricacy varied from pottery made over an open
reasonably controlled
fire, to the use of clever kilns where temperatures were
produced. Such a
were
enamels
slips,
glazes,
and in which ceramics with
medieval Nubia
in
operated
and
Faras
at
unearthed
was
ceramics
for
factory
(Adams 1977:496-8). Ceramic figures were usually made in kilns as well. The
products of work with clay were immensely varied from plain crude
the
kitchenware, to earthenware, faience, glazed wares, from plain bricks to
luminous 0a//i>- (Tunisian and other) tiles, from weights to figures, sometimes
of the
Glass
Sahara is the culture of
Africa
western
later
and
Northern
them.
to
allied
closely
was
manufacturing
made beads, while small sculptures or vases, bottles, and lamps were made in
and
Egypt. Painting on glass flourished in nineteenth-century Tunisia (Ayoub

life size

in the round.

known tradition of sculpture south

Nok. The range of techniques was vast.

The

oldest

Galley 1977).
Clay allowed the greatest versatility of expression of all the media used
sculpture. Ceramics could achieve intricate curved volumes as well as blocked
could
angular ones and could provide the finest detail. Hammered metal
effects.
cut-out
or
line
a
to
reduced
volume
achieve effects of silhouetting:

Ceramics could not achieve this. Perhaps metal was also a better medium to
intertwme space and volume. But even metal was not as versatile as ceramics.
The only limitation of ceramics was one of size: the size that could be managed
by the artist in the firing and the weight that made it moveable. Unlike metal,
uneven firing, variation in porosity, and even variation in the composition of
the prepared clays could all be exploited for effect.
Mud sculpture required little technical skill beyond the shaping of the
more
volumes on an armature. It was less versatile than ceramics because it was
to works which
fragile and had to be protected from humidity which limited it
them.
could be contained within a house or which could have a roof built over
much
details
and
massive
more
remain
to
had
It was also clumsier; volumes

worked out. Perhaps as a result, the product was often painted, much more
especially
so than ceramic sculpture. But mud sculpture was widespread

less

See Fagg and Picton 1970;

64

Coan and

Hauleville 1907.

MEDIA AND TECHNIQUES

known manifestations may be the mbari houses

and b, 1975), but tomb figures in the Zaire
1969a
Nigeria
(Cole
of southeastern
river bend (see Plate 8.5), large sculptures in eastern Angola, and initiation
statuettes from central and East Africa, ail deserve mention. Mud sculpture was
however an ephemeral medium. No nineteenth-century products even seem to

sub-Saharan Africa. The best

have survived.

SUBTRACTIVE SCULPTURE:

WOOD

In contrast to additive materials, subtractive media imposed more limitations,

but on the whole required fewer technical operations. The shape, texture, and
size of the product was largely determined by the block to be carved. The most
common materials in Africa were stone, wood, ivory and bone.

Techniques for carving wood and the tools used (adze, gouge, knife) were
simple but required great skill. A block was hewn out of a tree trunk, a branch,
fork or root. It was first proportioned for the main volumes, and then these

Plate 4.5 Carver. Bolony,

Kuba

as a person in the style of

some Kuba

country,

Ngongo ^roup. Photo June 1 956. He

dolls. Details

carves a paddle. Note the adze shaped

of costume typical of that penod except for the imported cloth

65

ART HISTORY

IN AFRICA

Plate 4.6 Seated figure. Wood. Use unknown. Kambundi,

Museum voor Midden Afrika,

Lemhwa Basuku, Kwango,

Zaire (Suku). Koninklijk

Tervuren. Height 27 cm. Acquired 1932. Shape inspired by forked branch used, or branch

chosen because the shape could be executed

were carved. Detail was then worked out, after which the wood was often
polished with sandpaper-like substances (often leaves) and could be given a
patina (Willett and Picton 1967; Willett 1978; Drewal 1980:9-18). The carver
had to choose his block with great care, visualizing the finished form in the
block. He had to be very attentive to the possibilities of the grain and often
66

MEDIA AND TECHNIQUES

worked patterns of texture into the general effect of the finished product.
South of the Sahara the most common block was cylindrical and finished work
either emphasized this, producing 'pole-like' sculpture, or, on the contrary,
ignored it so that other geometric volumes predominated. Sometimes forked
pieces or twisted roots were exploited, as for example in the country of
Bandiagara (Mali: Dogon) or on the Middle Kwango (Zaire: Suku) to produce
works in which the suggestive form of the raw material was emphasized. The
grain of the wood was an essential factor for the carving itself. Thus Lem
(1948) argued that Sahelian styles are angular and tall, with rather scanty
because the soft coarse-grained wood required this, while their southern
neighbours made compact, rounded sculpture, sometimes with exquisite
detail and almost lacquer patina, inspired by the possibilities of very dense and
fine-grained wood, which withstood such treatment. The finish, especially on
masks, included not only staining or a patina but involved polychroming as
well in the coastal lands between Togo and northern Angola.
If the effects of wood sculpture were greatly enhanced by inherent shapes
and texture, the size of such pieces was severely limited. Joinery could
overcome this handicap, but where joinery is found south of the Sahara, as in
the Niger Delta or in nearby Benin, the technique used shows European
influence. The relative absence of complex compositions in African wood
sculpture may be related to the monoxyle (one block) execution. Large
compositions could only be hewn out of very large trimks and were much more
difficult to envisage than if joinery had been common.
detail,

SUBTRACTIVE SCULPTURE: IVORY, STONE

Ivory, which was also widely in use all over Africa, was constrained by its
curved shape and its hollowness. It could be carved much more finely,
however, than wood, and was strong enough to be fashioned into a tracery of
forms barely tied together if this was the required effect. Most carvers avoided
this, however, and emphasized the solid volimie of the medium, made the best
use of the curvature and stained the product perhaps even more delicately than
stained. The fineness of the grain of ivory was thus fully exploited.
was
found only in Benin, but may antedate European influence. The
Joinery
best known example is that of the Bini leopards carved out of five pieces of
ivory, fitted one into the other.

wood was

Stone for sculpture in the round, relief or architectural sculpturing such

capitals or friezes had been used to perfection in north Africa, where
perhaps the highest technical achievements date from predynastic Egypt with
its sculptures and vessels in extremely hard stone as well as its translucent
alabaster. Working most stone was a matter of chiselling first, poUshing later,

as

on

except for soft stones such as soapstone (steatite) and some graphites which
were the favourite material south of the Sahara when stone was used. Such
media could be worked like wood and the stone sculpture from Guinea, Sierra
Leone, Zaire and Angola was clearly derivative of techniques used for wood.
Ancient Ethiopia, and North Africa before Islam, produced monumental
sculpture either out of a single block of material or by bonding several parts

when the intended size required it.

Sculptors in the north took advantage of every particularity of the

together,

67

ART HISTORY IN AFRICA

be of natural stain, roughness of grain, shine of polish or
marble vs. diorite). Their skill in the treatment of hard
greater than sculptors south of the Sahara who used this

medium, whether

it

reflectivity of light (e.g.

stone was

much

medium.

MIXED MEDIA
Mixed media were very common in the sculpture of Africa south of the Sahara.
The Bini ivory leopards have bronze spots and a Bini ivory mask wears iron
strips on its brow. Wooden statues or masks often carried additions of horns,
beards or dress, teeth, claws, bits of glass or of mirror. On the
Cross River, some masks were even totally covered with skin, thus hiding the
natural medium completely. Such adornments were often a requirement for
the use of an object, as the cavities covered by a mirror on charms in lower
Zaire, or the nails stuck in charms of the same area show, but foreign materials
were also often used to render eyes (shell, glass) or beards (raffia) and other
shells, fibre as

What mattered was the total effect. Masks were much more often in
mixed media than sculpture because they were only part of a dancing costume
- the total effect of a masked dancer had to be theatrical and often gaudy. Thus
sometimes the very qualities of the medium were lost when it served as a
support or carrier, rather than as the final form of the object in space by itself
attributes.

PAINTING

AND DRAWING

Because the graphic arts are two-dimensional, they require the feigning of
greater illusion than is required for three-dimensional arts: illusions of planes,
illusions of volume, illusions of space, illusions of texture, illusions of
atmosphere. Apart from the basic techniques, a range of technical skills to
create such illusions had to be added here (Gombrich 1960).
Basically, the technology required first the making or preparation of a
surface to carry the design or image: rock, wood, plastered wall, canvas, paper,
leather, even the skin of the human body. Most surfaces were then covered
with a ground for painting. The degree of absorption and the way in which
pigments or stains lay on the grain or were absorbed were crucial factors.
Pigments had to be manufactured for drawing, painting, or dying thread and a
certain amount of chemistry could be involved here. Hues were obtained from
organic or mineral products. The binding media that held the bonding agents
were mostly organic. Then the lines or paint or dots could be applied with reed,
pen, brush, or finger. Preliminary drawings could be executed for paintings or
engravings; it was common, for example, for Egyptian painters to outline the
composition first by placing red lines with a string dipped in colour. They
prepared such a detailed outline that each element could then be quickly
outlined and immediately given a ground hue, grey blue in the eighteenth
dynasty, later yellow, then white. Then the artist overpainted the flesh, the
clothes, the jewellery, the hair, using one or several layers of colouring as

required by the effect of opaqueness desired (Mekhitarian 1954:22-35).

The processes varied almost endlessly, not only for the two-dimensional
arts such as painting but also for other media such as tiles and potsherd
pavements. In painting, binding agents could be as different as encaustic (wax)

68

MEDIA AND TECHNIQUES

Plate 4.7

Tomb of Amen Khopshef, Valley of the Queens, Thehes. From left to nght: Thoth writing, Ramses III
I sis. Amen Khopshef. Hieroglyphs: above Thoth: Thoth records for you the renewal festivals of kingship';

embraced by

'

Great Divine (loddess'; to the nght: 'Lord of Appearances, Ramses,

justified. Lord of the Two Lands Ramses III'; under the Pnnce: 'Son of the King, of his Body, Whom he loves. Amen
Khopshej, justijied\ Tomb of Amen Khopshef, royal scnbe. Overseer of horses, son of Ramses III, Queen's valley,
between Ramses and his: 'To be said by

Isis, the

Thebes no. 55. VCallpainting, west wall, near entrance XXth dynasty c. 1180 BC. The painting technique is visible.
Gesso base on plaster, red basic drawing, dark colours painted first, lighter ones later. Overpaintmg on the headdress of
Isis. Hieroglyphs translated by courtesy of Prof M. Clagett. Photo Dr T. Webb
.

69

ART HISTORY

IN AFRICA

or glue boiled from animal skins and sinews, or the techniques could consist of
transferring patterns from a copybook (common in Coptic and Nubian wall
paintings) or apparently direct painting without any previous outline as in the
rock arts of the Sahara or the San. In other media, scenes obtained by the
juxtaposition of prepared

tiles

and the execution of mosaics were widely used

pavements in West Africa

in north Africa at different periods, while potsherd

reproduced geometrical patterns only.

The techniques of graphic art were more varied than was usual in
sculpture because of the special constraints of two-dimensionality Figures, for
instance, could be merely outlined, rendered by areas of colour only, shaded or
not to suggest volume, rendered by short strokes, zebra hatching, dots of equal
or unequal size, parallel and hatched strokes and so on. All of these techniques
.

found on Saharan painting (Lajoux 1977). Planes and depth could

be suggested by overlapping, sparing out, positioning on registers or

are already

foreshortening. In painting, the application of the paint for the final illusion
was frequently as idiosyncratic as a signature is; in drawing, the characteristics
of line were the equivalent of a signature. Because such features could be much

more individuahstic than those in carving, the study of such techniques may be
more revealing and rewarding to the art historian. The whole concept of ateliers
- workshops characteristic of style - and of individual hands, was first
developed in the examination of painting and such assignments often remain
most convincing for two-dimensional art.

TEXTILES

AND OTHER

FABRICS

decoration and painting of textiles, basketry and mats, as well as of other

such as barkcloth or leather or animal felt, is ubiquitous in Africa, and
has been for many centuries. In woven stuff, the art work is crafted in the
material itself during the process of weaving. First, the fibres must be made
ready from the raw materials such as raffia, cotton, wool or silk. The
technologies of spinning for the latter three materials vary, while for the first
with
the raffia strip is extracted directly from the foliole and then cut or carded

The

fabrics,

comb into thread. The thread can then be dyed if necessary either completely

dye
or tie-dyed (ikat: an Indonesian term) by tying threads together so that the
will not affect the spots where they are tied.
Weaving is done on a loom where the weft threads are inserted over and
under the supporting or warp threads. In much of Africa, looms have been the
or
only machines known in the precolonial period, and many, such as the Tio
Kuba weavers I have known, were quite conscious of this. The simplest looms
used were single heddle, the heddle being the device by which the threads of
weft
the warp which had been separated by a stick (the shed stick) so that the
passing
next
the
for
order
reverse
in
again
separated
are
passed,
thread can be
stick
of a weft thread. Double heddle looms used two heddles instead of a shed
and
pulley
heddle
a
and a heddle, linked them together with a rope going over
both
to
connected
pedal
foot
a
using
faster,
much
allowed the weaver to operate
heddles. Single heddled looms were all over the continent, while

double-heddled looms are found in the oikoumene and beyond in West Africa as
well

as

in

northeast

double-heddle loom

70

Africa.

that

The West

weaves narrow

African strip cloth loom

strips of cloth. It

seems

to

is

have come

MEDIA AND TECHNIQUES

PnT^ Corner stones and guard houses of the

walled and fenced palace property

Private quarters of the ruler and the

of the dynasty

03
//x\

|rx]|

^^"^

of the

paramount wives

Clubhouses

of palace officials

Huts

Queen-Mother

Audience chambers

of the ruler

r\
1/

Huts of

women

(wives,

relatives, servants)

Garden plots
X

Residences

Treasun/ huts

of officials divided

Shrines

into quarters

Granaries

Fig. 4.2 Plan of the

thread.

Bamum

Width 180cm

<

capital on textile. Cotton strip weave, indigo dyed. Stitch

length

540cm. Acquired 1936. After P. Gebauer 1979:374,

and dye with

pi.

resistant raffia

163.

71

ART HISTORY

IN

AFRICA

from the Nile valley in Sudan (Johnson 1977). Variations in the passing of weft
under and over warp threads yield different kinds of weaves from simple plain
weaves (once over, once under) or simple floating weaves (twill is twice over,
twice under) to complex compound weaves where supplementary wefts were
worked into the fabric, either over its whole width or over a part of it, often
with supplementary heddles. Different patterns can be created by the
alteration of the rules used to pass wefts through warp, and the patterns are
much enhanced by the use of coloured thread. The most complex technique
ever used in Africa was probably the work done on Coptic medallions and
borders, where up to twelve different dyed threads of wool were used.
Once the material is woven, it can be further ornamented by dyeing. The
most common technique was resist dye, where the cloth was tied or stitched so
that the dye would not affect certain areas. It also was possible to paint
freehand or to use a stencil to apply a resisting material to the cloth so that the
dye would not affect the parts thus covered. The contrary operation consists in
painting the cloth directly with dye, as painters do on a canvas. Patterns were
printed by stamps in Akan country (Ghana) and stencilling on woven stuffs or
fabrics is known from several areas. Applique, a technique found in the West
African coastal areas, in Sudan and in northern Nigeria, consisted of sewing
further materials on the finished cloth and thus produced figurative patterns,
letters or simple geometric effects. Other stuffs such as beads, shells, leather
or, in rare cases, metal could also be

sewed onto the basic fabric

to

produce

richly ornamented articles of clothing. Quilting and patchwork were less

widespread; the former was used onl> for armour in the regions around Lake
for winter coats in the Maghrib.
Embroidery with long threads was not uncommon and the production of
plush-like cloth is best known from central Africa. Here the embroidery thread
again, after having
is passed vertically through the fabric and brought back
been knotted or not, and the ends are then cut off above the surface of the
fabric to yield a pile. The Ulm cloth from Angola (Plate 9. 1) is an example of
this technique. Much rarer and from the same areas is open work, where, as in

Chad and

by different techniques, patterned gaps occur in the finished fabric.

technologies involved have best been described by Picton and Mack
Polakoff
(1979), but also by Lamb (1975), Loir (1935), Menzel (1972) and

lace but

The

(1980).

The history of the technology and decoration of textiles can be

reconstructed for the drier areas of the continent from surviving stuffs,
especially along the Nile valley. Iconography in European documents helps to
recover more data for West and Central Africa, while a few costumes and
materials from these areas were brought to Europe from the seventeenth
century onwards. But to date a history of textiles still remains to be written,
partly because they have not been recognized as art, partly because of the
technological history.
for making baskets and mats are similar to simple weaving,
but the plaiting is done by hand. Colours can be used here also and often
decorative effects were achieved either by plain plaiting or by the use of

neglect of

all

Techniques

contrasting colours (Coart 1927; Rossbach 1973).

72

MEDIA AND TECHNIQUES

ORNAMENTING THE BODY: SCARIFICATION AND PAINTING

body or tattooing have been mentioned by many authors,
but have in fact been the object of very little study. The technology varies: the
skin was cut and dyes could be introduced under it, in which case tattoos
resulted. Foreign bodies could also be put under the skin so that welts resulted
from the healing process producing keloids, a process usually called
scarification (Gebauer 1979: 104-5). Cutting and letting the cuts heal without
Scarification of the

any further operation resulted in cicatrization. All these interventions

produced permanent results. The processes were common south of the Sahara,
but not nearly as universal as is sometimes believed. Not all scarification,
tattooing or cicatrization was intended to be decorative - medicinal or magical
purposes and accidents produced scars as well.
Paint can be used to decorate parts of the body, such as hands, feet or the
face in a permanent way when the patterns are refreshed until the dye has
permeated the skin. Patterns of this type are found among Berber women, but
are much rarer than painting with henna or other substances for certain
occasions only. Tripolitanian women, for instance, paint their feet when they
are brides or attend weddings (C. Herman, personal communication).
Painting for initiations, rituals, war or festivals was very common all over
sub-Saharan Africa, although it is most often mentioned with regard to eastern
Africa. Sometimes the whole body was painted with, for example, camwood or

barwood. Often large portions were painted

decoration, or to frighten

enemy

in

contrasting colours as

warriors. Ritual painting very often covered a

limited surface and consisted of painting rings around the eyes or the mouth in
the basin of the Zaire river. The painting of decorative linear patterns was

known cases may be those of the country east of the

lower Niger and of northeastern Zaire, where the blue colour extracted from
the gardenia plant was used by women. The dyes most used were white or
yellow chalk, red from the camwood and barwood trees, and black from
vegetal sources, charcoal or soot. Usually these three colours and the painting
had a definite religious significance. It is noteworthy that not all dyes could be
relatively rare, but the best

used for painting the body or scarifying, as some were quite toxic; and that no
colour could be painted over too large a surface of skin for too long, as this
interferes with the breathing of the pores and can lead to severe illness or even
death.

and other matter, were common in

They have often been illustrated
(Paulme and Brosse 1956) but no in-depth study exists, yet they too were an
art. If some hairstyles were linked to ethnicity or with social position (as in
southern Angola), others were varied and developed mainly for aesthetic
purposes only. Some were creations for a few days only, but others lasted for
months and must be seen as more permanent art, not associated only with a
passing event, and in that they are comparable to masks.

Complex

many

hairstyles, sculpture of hair

west, central and east African regions.

MEDIA, TECHNIQUES

AND THE HISTORY OF ART

The first characteristic of interest to the art historian is the relative durability of
a medium. Thus ceramics and stone last almost forever, basketry, matting,
73

ART HISTORY
and

IN AFRICA

mud the least, and many claim that wood under tropical conditions rarely

a century of age. Be that as it may, it is obvious that our knowledge of

forms in the past is heavily influenced by what has survived, and what can
still be found in the ground. This means that our chances for arriving at a
general history of art objects in metal is good, with the exception of gold which
is melted down too often, but for objects in mud, it is worst of all. In
sub-Saharan Africa, works of metal, ceramics and ivory, all long-lasting
media, are more numerous than was once believed, and the prognosis is good,
but can we reconstruct general features even of sculpture from works in
ceramics, ivory or stone? Are these media so different from others, especially
wood, that we will never be able to trace even the main characteristics of the
bulk of sculpture here? Perhaps not, when we consider works in metal and
ivory. Certainly, as the comparison between a Kuba work in iron and one in
wood shows, there are some resemblances, but the execution is altogether too

exceeds
art

Even ivory, except for small pieces, like masks of Maniema (Zaire)
do not enhance the natural curve of the medium, may present differences
that are too great, but when ceramics, stone and wood are compared, the
different.

that

prospects can be better. After all, clay can exhibit technical features
(angularity, flat planes) which are more natural in a wood medium but are also
easy to achieve in clay, and stone sculpture in steatite or graphite was worked
like wood. So media in clay or soft stone, I believe, can often be compared to

wood.

Sometimes, however, different media are handled by different

conventions. Willett (1971 ill. 176-8) gives a striking example of a naturalistic

rendering in pottery, as opposed to stylization in wood, in works from a single
society! The reader can easily explore the problems of comparability by
examining the body of sculpture in western Nigeria where work in all the
media mentioned has coexisted. The transition between the stone figures of
Esie and presumably wood sculpture is easy, it is also easy between ceramic
heads and brass or bronze work at Ife, but nineteenth century wood sculpture
:

and brass (Ogboni) statuettes are not easily comparable. Work in ivory also
differs considerably from comparable objects in wood. On balance, then,
reconstructions of the past will never be full due to losses through
impermanence, but main traits may be determined to some extent.
A history of drawing and painting will be even more difficult to recover in
many parts of the continent than sculpture. Often the supports were in
perishable materials such as mud walls of houses, bodies, or clay platforms.
For textiles, the conditions of preservation are very good in dry areas but not
elsewhere - Nubia may have the longest sequence in the world of preserved
textiles. The reader will note that styles of textiles and styles of body decoration
are often quite comparable, so that the former can shed some light on the latter.

The next point in the consideration of techniques is that questions as to

chronology and even place of manufacture can be much more complex than
first imagined. What we see now can be the end product of a very long history,
and this is particularly obvious in complex architecture. But even other works
made within a workshop pose problems of authorship: was it the same hand
and that finished the details of a wooden sculpture? Was it
same hand that worked the main volumes by chipping with the adze? Was
it that same hand that blocked out the main volumes? Often not, and often a
statue is the product of a workshop group rather than of a single master.

that applied patina

that

74

MEDIA AND TECHNIQUES

Questions about

when and where become more complex when

the process of

the manufacture of the prepared medium is concerned. Questions as to the

preparation of the medium itself may not be directly relevant, and fade into a
general history of technology, but they cannot be avoided! Where did the iron

Where did Benin's copper

bronzes?
or the alloy used in the Igbo
Also unavoidable are questions of technological development. How was
West Africa
lost wax actually produced? We learn that it was not indigenous to
but reached Igbo Ukwu perhaps by the ninth century A.D. (Shaw, 1970a,
1970b, 1978:118-19). Its distribution in West Africa is consonant and
explicable in relation to trans-Saharan trading routes and other channels for
diffusion in the area. Close examination of the use of the technique in southern

come from
come from

that

was used

in the

Kuba

statuette?

Ukwu

Nigeria shows that instead of wax, latex from the euphorbia was used. But the
euphorbia does not grow there; it only occurs further north. Details of Igbo
Ukwu pieces convincingly show that latex was used there, too. So before the
ninth century, lost wax must have been used in northern Nigeria and long
enough to allow for the development oi ihe euphorbia latex technique (Williams

1974:179-2!3).
Questions of technology help to link styles together as well as techniques.
The fact that goldsmiths in Ghana imitate fihgree, proves that earlier foreign
work was
filigree work must have fascinated them (Garrard 1980:108). Such
current in northern Africa. Once we know this and additionally take into
account that north African vessels have been found in the area, there is little
doubt that comparable decorative work done in filigree or by the Ghanaian

Fig. 4.3

/.OS/

wax

techniques in West Africa

75

ART HISTORY
method

is

IN AFRICA

similar because the smiths there copied the result of filigree

working.

Skeuomorph tendencies are also historical clues. The hypogee churches

hewn out of the living rock reproduce all the elements of a

of medieval Ethiopia

basilica, all functionally useless here, such as beams, lintels, arches, corbels on
columns and domes, even the upper clerestory windows as blind tracery. They
prove that basilicas built in the open preceded them. They also raise the
question of possible ancestry for an architectural vision able to plan a whole
church from a single block by excavation. Was that new or not? In fact it was
not. Was it an Indian inspiration as some thought? It was not. The presence in

^vi;?c
!%r^-^

Plate 4.8 Fagade.

Ethiopia.

9m

by

window frames

76

Abba Lebanos church. Carved from the live rock, which still forms the roof. Lalibela, second cluster.
7m by 7m, almost a cube. .4.D. 1190-1225. Three nave basilical plan. .Axumite elements e.g. in

MEDIA AND TECHNIQUES

hewn catacombs at Axum as well as the carving of rooms
underground at Matara in the early centuries of our era, if not before, tell us
that. The rockhewn churches were a unique combination of two traditions that
had coexisted close to a millennium in Ethiopia. Their fusion produced that
astonishing phenomenon: the basilica concept rendered as a piece of sculpture
Ethiopia of rock

(Gerster 1968; Chittick 1974, 1976).

The examination of the precise medium and the technology used in the
production of a work of art is indispensable to the art historian. It may help him
or at the least establish possibilities. Lost-wax work in lower Zaire implies that
the technique
feat, or that

it

was either invented there, a remarkable - almost unbelievable was imported, probably from West Africa (Wannyn 1961). But

when and how?

It

allows historians to place a

work

in a technological series.

Admittedly this is much more common in architecture than for the other arts
and then especially with regard to Islamic, Ethiopian or classical architecture.
Often technical innovation and new forms go hand in hand. The development
of the stalactite (muqamas) from its simplest form in the tenth century Qala'a of
the Banu Hammad in Algeria to its triumphant transformation of the surfaces
in the honeycombed domes of the almoravid mosque Qarawiyyin in Fes is a
story of stylistic evolution as much as one of technical development (Margais
1954).

Thus

the historian

take place.

Why

is

relationship to the

is

led to ask

how

or

why

certain technical innovations

North Africa? Did

the dome so important

domed structures of the Saharan
in

it

have a

populations, reported

since Garamantic times before our era (Daniels 1970:41-2)?

Or must we

the Middle East?

Sahel, as well as
in
the
and
Africa,
northern
Probably both the use of domes in
Near-Eastern techniques, have been involved.

believe those historians

who

derive

all

domed forms from

In their beginnings, most techniques have little to do with art. They were
invented or adopted for purely practical purposes. Thus raw materials were
transformed into media because cooking pots or iron tools or cloth were
needed, not to provide means for the artist. Artists usually adapted what they
found. But beyond this, certain techniques, such as the lost-wax process or the

construction of domes, were not directly utilitarian. They relate to art and it is
to them that the art historian must give his closest attention. Sometimes, and
with this example I conclude, both artistic and utilitarian purposes become
blurred. The Copts developed a highly sophisticated technology of weaving,
allowing them to create inimitable masterpieces of art in textiles (Grigorieva
1980:28-9).
the

The technology was

Roman Empire.

by the demand for such products in

Kuba developed more dyes for textiles than

fuelled

Similarly, the

any of their neighbours and kept the techniques secret, for their polychrome
cut-pile textiles were their major export in the eighteenth century, as they are
again today. However, once a medium was available and a technique was
developed, artists used these materials and procedures, and so the history of
technologies

become

part of the history of art.

77

CHAPTER FIVE

STYLE
The study of changes in shape over time is held by many to be the core of art
history. Once any object, such as the board from Ardra, has been locahzed and
dated, once

its

context

is

known,

it

can be put in a

set

of series of similar

objects: series of technically similar works, series of objects similar in subject

matter and a series by shape or style.

Style refers to the formal properties of works of art (Layton 1981:134;
Focillon 1943: 1 1). The concept designates the shapes of each component in a
given work of art, as well as the composition of all the elements into one overall
visual and tactile design. The term can be applied to the formal properties of a
body of art works as well as to a single work. Style, as the summation of design
choice, may have different characteristics in the works of each individual
artist, in the arts of one locality as compared to another, or as changing tastes in
design may occur as Feriod follows period. Therefore studies of the style found

and of a period in time, are major means by

which an ordering may be established in the history of an art. As style excludes
meaning, a stylistic history by itself cannot achieve understanding and hence
cannot be art history, even though it is a core component of such a history.
The relationship between art objects and the things they represent is far
from simple. Scholars still dispute how exactly a natural thing is related to its
representation in art. Gombrich has conclusively shown (1960) that their
apparent sameness is an illusion. A representational work of art is a natural
object interpreted and simplified. Representation always involves a reduction

in the arts of a place, of a people,

all the features of natural reality into a few that are significant to the artist
and the community involved. The reduction uses schematic patterns,
conventions to realize this reduction. Such stereotypes are the language of
form (Layton 1981:144, 161-71). All scholars agree that there are always

of

stereotypes, there is always convention. Style then refers to the sum total of
such conventions in a body of works of art or in a single piece. It refers to
formal elements common to a series of works by one or several artists and also
to formal elements that are uncommon or even unique. For a convention can
be unique, as long as it is understood.

CONVENTIONS
Let us look

first at

an ancient Egyptian painting. Certainly

its

arbitrary, registers separate scenes, the sides of a scene are laid

perspective is
open like the

of an opened box (Ragghianti-Collobi 1968:52). Odd things happen to

is frontal for the shoulders, torso and eye, but in profile
for the other parts of the body. The depth of the field and planes are indicated

lids

figures: the perspective

78

STYLE

surmounted by heron. Wood. Face black and bnlhant patina. Hair, black, white, red. To be
and meaning unknown. Central Ivory Coast (Guro). Museum,
University of Pennsylvania. Height 52-6cm. Acquired 1929 from a US collector. Sharp-nosed style; perhaps, hence

Plate 5.1

Homed mask,

held in front of face by teeth of the wearer. Exact use

from Zounoula area

79

ART HISTORY

IN AFRICA

hones. Taki. Museo

Plate 5.2 Rthquaiy Jinure. Wood, copper, hras^i. Used on lop of the bundle holding ancestral
head oj the villane
premonco cd eino^rajuo, Rome. Hetuht 42 7cm. Bought hy A. Pccile, Sept. II, IHH^jrom Taki,

near l-'ranceville, then inhabited hv

80

Ndumu. See

h'lu. 5. /

STYLE
by the size of figures; the colours are largely
look upon such paintings with an eye trained by
post-Renaissance conventions in Europe, all this hurts and we cannot relate to
Planes are not integrated, perspective is not unified nor is light,
it.
foreshortening is unknown or almost unknown, composition is based on the

by overlapping
conventional.

figures, but not

If

we

must be read in one direction only, and does not conform to the clever
geometric figures in European art. Modelling to render volume by shading of
colour is also absent. We come away from the contemplation of an ancient
Egyptian painting convinced that all styles are conventional and create
illusions of reality by the mutual understanding of how to read images that
exist between viewers and makers and that such and such conventions stand
linear,

such and such realities (Gombrich 1960).

Representations in two dimensions are more dependent upon synoptic
illusion than are three-dimensional representations, while carving, because it
is subtractive, may be more in need of stereotype than additive sculpture. But
whatever the technique and medium, renderings remain arbitrary to a very
large degree. Thus the classical sculptured Greek nose has its counterpart in
the upturned nose of one Ivory Coast sub-style (Guro, Plate 5.1), or the single

for

accolade of the eyebrows in the Kwango-Kasai area (Pende). None of these is

really close to reality as perceived. It is easier to realize that convention rules in
very styhzed or expressionistic art than in 'naturalistic' or idealized art; but in
fact conventions are almost as strong in one as in the other. The reliquary
figures of eastern

Gabon whose heads become two-dimensional

or almost so

where torso, arms and legs are caught at a

glance, except that there is no torso below the neck and the legs are in fact a
simple columnar volume (Plate 5.2), still, we see that these are people and we
realize that this type of realization rests on a collective convention. Yet we

and whose bodies are

just a lozenge,

should also realize it when we look at an extreme in naturalism such as Ife

figures or headpieces from the Middle Cross River (Ekoi). The ears, the eyes,
the lay of the symmetry, the mouth, cheekbones, etc., are all 'artificial' on

Gabon. This is a drawing of the item plate 5.2. Copy of A. Pecile's

drawing of 1883 (H. Zorzi, Al Congo con Brazza, Rome 1940:312). The drawing establishes the identity of the piece.
Note differences m medium between drawing end photography
Fig. 5.1 Reliquary. Taki, near Franceville,

81

ART HISTORY

AFRICA

IN

closer inspection; they are not like 'real' ears, eyes, mouths, cheeks.
composition is as conventional as the Gabon reliquary.

The

Even in architecture convention rules. Intended use certainly prescribes

the size and sometimes the general plan of a shelter. But the buildings can be
high, low, roofed in any of a dozen fashions, wider or narrower. Traditions in
the realization are as strong here as in any other art form. The conventions
ruling ancient Egyptian temple building lasted for just as long as those
pertaining to hieroglyphs. The basilica, an imperial Roman achievement,
became the model for Christian churches and remained one of the models for

churches over the ages, being virtually unchallenged

The

Ethiopia.

basic features of the

ground plan of

Nubia, Egypt and

in

a classical

mosque

as

today and
plans of
round
The
mosques.
of
styles
the
other
of
just as arbitrary as any
Zimbabwe, the rectangular shapes and their assemblage around quadrangles

opposed

to later

Mamluk, Libyan

or

Ottoman plans

are

still

alive

of palaces in western Nigeria, or the basic features of a palace on the Cameroon

grasslands with their mixtures of pointed beehives on square bases, all are
equally arbitrary, and equally conventional.
Such stylistic conventions are learned by the eye as objects come to be
perceived as visual concepts. A mosque to any Muslim is only one of a few
models of buildings. To the Cameroon grasslanders, a palace must have
beehive doming on square bases. Artists learn their conventions along with
their craft. Apprentices imitate, as we have seen, and this includes all the
recipes for rendering all the varieties of reality that are rendered in that
workshop. Moreover, conventions da not operate one by one nor are they

grammar: everything hangs together to a large

and a copy of a single subject from one style into
another becomes a total translation, as Leroy (1967:57) has so beautifully
shown by reproducing side by side an engraving by Tempesta (1591) and an
Ethiopian replica. The differences in the works do not concern this or that
detail - the eyes, for example - but everything. To Europeans the Ethiopian
learned as such. Style

extent by

common

is

like

rules,

replica looks vaguely like a caricature only because that

is

the effect of the

systematic application of foreign canons of convention.

Style then

in a given work and in series of works from the

phenomenon. Often all works in a given medium

everywhere

is

same workshop.
exhibit the same

It is

a total

basic style for long periods in a given cultural continuum: the

Ethiopian style in painting lasted over a millennium. Within that framework
there will be stylistic differences just as the speech of one person differs from
that of his neighbour, but the language remains the same.
The conventions of style are arbitrary and hence eminently rooted in time
and place, but not totally arbitrary, because style is constrained by intended

use and even more by the medium and technology available. Thus, a stone
house on the coast of Kenya in the eighteenth century had to have its entrance
and all other doorways located in such a way in relation to each other that
complete intimacy of the inhabitants, especially the women of the house, was
preserved (see Fig. 11 1 and De Vere Allen and Wilson 1979: 1 1). Their rooms
were long and narrow because the mangrove poles used as beams to span the
width are never very long. That characteristic helps to explain a stylistic
element unique to those houses: the development of multiple niches in the
.

walls, of varying

82

depth and

slant so that the illusion of greater distance

was

STYLE
given to the roving eye, than the actual distance. The room looked much
roomier and the wall further away than it was.
Such constraints are obvious. Masks have to be carved in light wood as
they are worn during dances, and they must have slits somewhere to see
through. Charm statues in lower Zaire or in East Kasai (Songye) must have
hollow bellies or horns on top of the head to stuff with medicine. The size of a
raffia cloth was set by the length of the fibre extracted from the palm frond.
The size of a support determined the size of any two-dimensional work.
Frescoes had to be executed very quickly, because of the technique involved,
so patterns were prepared in advance and applied to one section of wall after
the other, and in turn that led to striking repetitions with minor variations as
the same stencil was used over and over. Once patterns were evolved and
copied, they could be transferred to work on other surfaces such as the canvas
used as support in Ethiopian churches after c. A.D. 1400.
As we have seen, techniques changed and with them the constraints of
conventions changed. But the constraints imposed by the media remained
much more constant. The fluid line of the cast metal could not be achieved in
stone or wood, while the special density of angular volumes balanced against
each other in wood could not possibly, in Africa at least, be executed in iron.
Nevertheless skeuomorphism occurred and involved a partial translation of
the style appropriate to one medium into another. The incised calabash
imitated in wood keeps its general shape, albeit more regular than most
calabashes would be, and the fine incisions turn into deeper and angular
engravings. In Ghana, the theme of the mother and child on a chair may have
first been made in ceramics, and replicas in wood became much more precise in
line and set the volumes in sharper contrast (De Grunne 1980:152, 155).

But the constraints of media never wholly dictated style. The most
its fundamentally arbitrary character
which gives each period and each culture its uniqueness. Labels such as
striking feature about style remains

appended to
by Europeans for that is what they remind European
viewers of, not to mention the use of 'archaic', 'classical' or 'baroque' and
indeed 'rococo'. The use of such terms highlights the pitfalls of stylistic
'cubist' or 'stylized' or 'idealistic' or 'naturalistic' are often

different African styles

assignments. First they are obviously ethnocentric; secondly, quite a number

among them imply general theories relating to the change of shape over time
whether that be expressed in the series 'archaic' etc., or by more
innocent-looking terms such as 'abstract' or 'naturalistic'. Finally, such
practices highlight the danger of

summing-up

stylistic

characteristics in

and impressionistic terms. Just as one can discover the

grammar of a language one can find the components of style in an analytical
way. In African art south of the Sahara, this approach has been called
essentially holistic

morphological analysis.

MORPHOLOGICAL ANALYSIS
The

goal of morphological analysis is to establish to what degree a style differs

from all neighbouring styles, the main principle being the arbitrary character
of any representation. If representations resemble each other in the whole
composition as well as in a number of dissociated elements, they must be

83

ART HISTORY

IN AFRICA

historically related, usually as replications

from

common prime work with or

without further borrowing.

F. Olbrechts (1959:29-35) was the first scholar of Africa south of the
Sahara to develop the techniques by which styles can be established. Four
levels were distinguished: the element of added and decorative detail, the
position in space, the proportions, and the sculptural detail. Added detail
involves, for instance, representation of hairstyles and tattooing, while
sculptural detail deals with the

way

dissociated elements such as the nose, ear

mouth, have been rendered. Added detail is very useful in localizing objects,
since emblems, costume, hairstyles, scarifications, jewellery, reproduce those
used locally at a given time. Sculptural detail relates more directly to core

or

conventions. Position in space concerns the linking of different volumes, for

instance, the transition between torso and pelvis, the main axis (repetition of
volumes, accenting, etc.). Proportions concern well-known canons of
sculpture.

height in

For instance, the human head

many

is one-third to one-fourth of the total

African sculptures south of the Sahara, while the proportion is

seven or eight in European art.

art, a study of position in space or a study of
proportions will be dominant. In the analysis of the art of the Bandiagara cliffs
(Mali: Dogon) the first is crucial, whilst the latter is important for the ancestral
figures of northern Gabon (Fang). In any case, propositions derived from the
examination of one of these four categories are much strengthened by the
eventual concordance of the results from the examination of the other
categories. A study of the art of northern Gabon illustrates the technique.
First, L. Perrois (1972) assembled artifacts from a known area (Fang) into a

one

to

Depending on the

corpus of alike-looking objects. Two to three hundred ancestral statues seem

sufficient for an analysis where similarities are great, while a larger number is
needed for more variable images such as masks. The numbers must be great
enough to achieve statistical significance. So Perrois measured proportions
such as the height of the head in relation to the height of the torso and that of
the lower limbs. He found that his statues were reduced to four groups only:
hyperlongiform (very long - comparing head to torso) torsos, longiform (long)
torsos, equiform (equal length) torsos, and breviform (short) torsos. As carvers
begin by establishing the relative proportions of the head to the torso to the
lower limbs, the measure was objective. He then turned to the position in
space, classifying general head shapes and the relationships of volumes. These
are obviously in large part determined by the canon of proportions used. Then

He distinguished, for instance, between

no fewer than seven types of eyes: mirror applied with resin, coffeebean shape,
copper plate, incision, disk shape in relief, rectangular shape in relief and
Perrois turned to the associated detail.

absence of eye indication.

Every single object fell into different groups for differing measurement
but the overriding classification - here the proportions - was the one where
covariation of the greatest

number

of features coexisted, showing objectively

the greatest internal consistency. Last of

holding horns or not, was considered.

the following criteria in rank order:

84

all,

The final

added

detail

such as hands

Perrois classification included

STYLE
1.

General proportions:

torso/ whole height

Position in space:

neck/whole height
relative concaveness of head

head/legs
2.

frontal/lateral proportions of the

head

Style of hairdo

Position of arms

(attached to sides or not)

Position of legs
3.

Associated details:

types of nose

mouth: shape, with or without beard

ear:

shape

(5 types)

eyes: shape (7 types)

navel: shape (4 types)

breasts: (4 male, 4 female types)

could then be described in the relevant terms such as: 'Longiform.

40 to 50% of the height, legs 10 to 30% of
the height. The torso is half of the total height. Hands to the belly, with mouth
at a level with the chin or hands holding a horn, associated with a bearded
mouth'. Only the general shape of the head did not vary along with the other
criteria. Ornamentation personalized objects. It made 'portraits' out of the
ancestors represented by the treatment of hairstyle, tattoos, jewellery,
eventual dress and the objects held in the hand.
The validation of the classification was reinforced by the coincidence of
stylistic variation with spatial distribution, linking the stylistic categories to
practices in different workshops, although Perrois expressed them in terms of
ethnic units. The basic division was between northern hyperlongiform and
longiform styles called Ntumu and Ngumba and southern equiform to
breviform styles called Fang, sub-divided into Nzaman/Betsi, Okak/Mekeny
and the extreme breviform Mvae. From that he could conclude that there had
been two centres at first, because there were two styles: one north (Ntumu),
one south (Betsi), although it is not evident why he did not also consider the
Mvae as a centre, since there is little overlap between them and other southern
styles. The ethnic labelling introduced once again an element of fuzziness. For
Ngumba do not belong at all to a Fang cultural and especially linguistic
grouping and the use of other names begs historical questions, as to ethnic
identify, relative position of different groups at different times (the pieces in
the corpus are not dated and could vary considerably in age) and so on.
The end result of this example of morphological analysis was firstly the
more objective description of the whole corpus. All the works had concave
faces, high and swelling foreheads, hairstyles in the shape of a helmet, round
body volumes, prominent muscular relief of arms and calves but less so for
shoulders, torso and thighs, bottle-shaped torsos and a general stance of
immobility. The whole corpus was called Fang. Secondly, the substyles
distinguished could be put in a series of variations. The analysis established a
stylistic series over areas, which presumably corresponded to a time series and
a succession in time of 'prime works', works showing major innovations.
Morphological analysis proceeds differently according to the corpus and
its characteristics. Thus, for Liberian masks, the first basic distinction made

Every

style

Head 20

to 30"^ of the height, torso

85

ART HISTORY

IN AFRICA

Plate S.i Ancestor figure. Wood, feathers, harkdoth. Left part offace and left leg white right part offace red. Usedon
Gabon (Fang). Museum fur Volkerkunde, Rerlin-Dahlem. Height
,

top of a box containing ancestral hones. Northern

65cm. Acquired 1882. Longiform

86

Perrois' morphological classification

STYLE

.24

36

.57

3 3
.1

Fig. 5.2 Morphological analysis of Fang (Gabon) statues, after

the length oj the torso in proportion to that

traditionelle

of head and

legs.

Perrois. Hyperlongiform and breviform according to

Hyper-: very long; hrevt-: short. After La sculpture

I.

du Gabon, Pans 1977:33

was between 'expressionism' as opposed to the degree of 'idealistic realism'

(Vandenhoute 1948). In F. Neyt's recent exhaustive study (1977: esp. 48-50)
of ancestral figures from northeastern Shaba (Hemba), Perrois' strictly
statistical approach was used only as a general but perhaps secondary
technique. Selections were first made between prime works or prototypes and
replications or copies. Neyt distinguished eleven styles and atypical works
based on the shapes of heads, eyes, nose, mouth, ears, neck, the plane of the
shoulders, the torso from four vantage points, the arms, hands, pelvis, feet and
stand. The secondary characteristics of hairstyles, finery, scarification and
prestige emblems were added to this. By then considering for each of the styles
what the prime works were in relation to the replications and comparing them
to those of the other styles,

he attempted to establish an hypothesis about

development. Despite the lesser reliance on statistics, the attempt
was in fact stricter than that of Perrois, because it forced the author to consider
each step from prime work to replication and from workshop to workshop.
The main danger of morphological analysis and taxonomizing remains
subjectivity. The units chosen for measurement must correspond to those that
were considered significant by the carvers and their public. Where field work is
possible, errors of this sort can be avoided by studying the practice and the
aesthetics of workshops and artists. Otherwise, one must make assumptions
through analogy. More problematic is the impossibility of measuring
everything, including some of the most significant signs of individual
signature. The stroke in painting, or the 'attack' of an adze in wood sculpture
historical

87

Plate 5.4 Ancestor figure. Wood, leather

'Urua'

I.e.

Hemba

(Shaba).

Museum

belt, raffia skirl.

Berlin-Dahlem.

Height 81cm. Acquired 1897. Neyt 1977: 'classual

Niemba

style'

STYLE
cannot easily be integrated into a system of measurements, nor can more
obvious visual clues easily be quantified. Thus, a given line for the bridge of a
nose is there, or it is not, or it is 'sort of there. All that can be done
comparatively is to organize such lines into one or several categories and then

them

directly as a criterion. Statistics become less valid here than in the

of proportions and the morphological outline tends to become a
comparative shorthand description. As long as this is realized and as long as it

use

case

is also realized that in setting up categories of such phenomena arbitrary

decisions necessarily intervene, morphological analysis still has the great
advantage of making comparative intuition explicit, and replicable by others.

Subjectivity remains but its analytical progress can be traced. Nevertheless,

morphological analysis is only an aid to better informed qualitative assessment.

ATELIER AND ATTRIBUTION

The more
disclose

refined morphological comparisons are, the more styles they

in the previous example for Shaba, an area that was about

- eleven

100km by 100km. The tiny Suku group in Kwango is credited with four
independent styles, their neighbours, the Pende of Kwango and Middle Kasai,
with three (Neyt 1977:430). The eleven Hemba styles were crafted in thirteen
administratively recognized chiefdoms, grouping ten villages or less, perhaps
each style corresponds to a dozen settlements. We are close to the level of the
workshop, and so we should be, for clearly styles are transmitted in workshops
and workshop traditions should be discernible in their output. We will call
them

To

ateliers to draw attention to their role as a unit of tradition in art history.

find the atelier that corresponds to a style is a major discovery.
By the 1930s Olbrechts (1959:71-5) had recognized an atelier in

southeastern Zaire which he named Buli, after the provenance given for some
of the wooden statues which shared the stylistic characteristics of the

name was an administrative designation. The

however, was so similar that to this day, some defend the notion that all
Buli works were the creations of a single artist, called 'the master of Buli'.
Recently one statue was clearly traced to the village of Kankunde, chiefdom
'long-faced style of Buli'. But the

style,

Nkuvu, to the northeast of Kongolo in eastern Zaire (Neyt 1977:321, 442).

The statue was so famous there that the genealogy of its keepers was known for
four long generations, with several keepers per generation, back to the first
forty years of the nineteenth century. The statue was carved by a certain

Ngongo ya Chintu, who

stylistic characteristics

lived at the village of Kateba, close by. Better known

of that area allow us to say that the Buli atelier shows

evidence of a triple

stylistic input; from the area itself, from the southwest

(Luba) and from the northwest across the Zaire river (Kusu). If there was but a
single master of the style it was Ngongo ya Chintu. But there were probably
several. We still do not know whether the atelier lasted only a single
generation, early in the 1800s, or several. Perhaps further inquiry can settle at

least the issue of the number of masters involved. The conjunction of historical
data and of stylistic analyses has allowed the art historian to account for the

style

and suggest the general conditions of

Because

classification.

style

One

its appearance.
linked to an atelier there is a spatial aspect to style
should expect therefore that transitional styles could occur
is

89

ART HISTORY
between

IN AFRICA

stylistic centres.

central style
typical style

This

is

precisely

what happens. Indeed

own

in

some

cases

Consider the typically

of Shaba (Luba) all built up in rounded volumes and then the
of northeastern Kasai (Songye) with their angular blocked out

the transitional style

is

quite a feat in

its

right.

By Ngongoya Chtniu? (Rult Workshop). Vi'ood. Used hy diviners. The bowl may have
woman represents a nature sptni mutitenta. Kateha, Shaba, Zaire. Konmkhjk Museum voor
is
Tervuren. Height 53 Scm. Before 1840. A product of the 'long face style workshop' of Bull. Xgongo

Plate 5.5 Kabila yi^ure.

contained while clay. The

Midden

Afrika,

the only carver

90

of

this

shop known hy name and

may have carved

all

works

in this style.

Dated hy genealogy

STYLE
geometrical volumes, almost cubistic. A transitional style is hard to imagine
here. And yet it did exist and yielded some very striking masterpieces.
Historically we do not know where the ateliers were located, nor when. They
must have been somewhere on the border of the Luba empire with the Songye

Plate 5.6 (.:hurm figure. Wood, cownes (eyes), oiled. Use unknown. Hast Kasai, /.aire. Htnograftsh Museum,
Helf-ium 1920. The comhmalwn oj very angular and rounded slyle suggests an

.Antwerpen. Height 40 5cm. .Acquired

ongin among the southern Songye chiejdoms bordering on the I.uha empire

91

ART HISTORY

IN

AFRICA

chiefdoms. But we do know of other strong influences from the Luba empire
on certain of these chiefdoms (Fairley 1978). Moreover, the chiefdoms were
organized in towns and the numbers of possible localities is restricted. The
general area and the general conditions of cultural influence are now known
and such transitional styles between Luba and Songye are placed.
Significant distributions of style over space are not limited to sub-Saharan

A striking example is the architecture of domed tombs in the Sahara

and its northern fringe from the Atlantic to Libya and its extension into eastern
Sudan and the Horn. Thus Margais (1954:435-7) distinguished five types by
area, three of which were found from the Horn to the Atlantic. The two others
show Turkish (post- 1500) and Andalusian influence. One of the widespread
types is linked to the structure of the oldest dome at Qairawan, but we still do
not clearly know which came first, the tomb or the mosque! Similarly,
so-called pillar tombs of East Africa and Madagascar fall into few styles.
T. Wilson enumerates eight main types of East African monumental tombs. The
relationship, especially where stepped corners exist, between such tombs and
those of northern Africa, is evident, but the whole question, like that of the
Saharan tombs itself, still awaits its investigation.
To link style and atelier, indeed even a master, is not equally possible with
works in any medium. It should be easiest in painting, for the very notion of
atelier developed by studying style and 'hand' in European painting. It is quite
possible in carving, especially perhaps in wood, but it is much harder in
additive sculpture where the plasticity of terra cotta and metal makes
replication much more nearly perfect on the one hand and allows variation in so
many directions on the other, that no corpus ever seems to include enough
pieces of the same 'style'. Differences in individual works by the same hand
tend to be taken as different styles! The technique has not been widely used
with regard to northern Africa, yet it should be able to extend our knowledge
with regard to such different arts as painting in Ethiopia, the smaller rock
churches there, small-scale architecture in northern Africa, works in painted
ceramics, textiles and perhaps decorative woodwork.
To achieve the recognition of an atelier or of a single master by the
examination of their products is remarkable, but morphological and other
Africa.

examinations alone can only lead to probabilities. We know well

in European art that while some have stood up
against later discoveries of data, others have not. Nevertheless, the attempt to
link style and atelier is crucial to the endeavour of the art historian and should
not be abandoned. Statements about it must however be capable of objective
inspection. To invoke the invidious distinction between those who have the
aesthetic eye and can see intuitively that these two works are by the same hand,
and those other unfortunates who do not have an aesthetic eye, is simply
stylistic

enough from such attempts

nonsense,

because

scholarship.

replication

and result is essential to

and other stylistic shorthand

of reasoning

Thus morphological

analysis

techniques for comparison are of the greatest value.

STYLISTIC SERIATION
Morphological analysis usually leads to seriation. This is the practice of placing
the objects studied on an imagined continuum of forms, whose poles are the

92

STYLE

denve from general architectural style and are a good

c; R Turkish influenced style, Aigena, since the 1 7th c^; C
south of the other styles and A lie
Andalu^ian style. Morocco and western Algeria, since the I2ih c; D Algerian plateau,
consists of a semidome on a cube, and occurs in
valley since c. 8th c.C) linked to nomads. The simplest style (not shown)
varuints occur m the Sudan and the Horn oj
easter^ Algeria and Tnpolitania as well as elsewhere sporadically. Other
Fig 5 3

Tomh

of

Hoh

Men. Sorihem Mashnh. The

indicator of cultural influences.

.\

Tunisian

styles

style since the 9th

Africa

93

ART HISTORY

IN AFRICA

SAMadonna. Detail. Fresco' Bishop Mananos, protected by Madonna and ChnstchiW in South Chapel, eastern
wall, third layer of the Cathedral at haras, Sudan. Face 18cm high. Polychrome period, after A.D. 1005. The round
face of the madonna, her double chin and the treatment of the skinfolds of the neck are typical for the Master of the
Madonnas, who was active from this date to almost 1 1 00. A nother master painted the Chnsi figure and yet a third bishop
Fig.

Mananos, a remarkable

94

division of labour

STYLE
most extreme

variants. Seriation

is

implied in the Fang corpus or the Hemba's

eleven styles. Seriation implies time, either quite consciously and explicitly, as
in the Neyt approach, or by the less explicit approach of Perrois.
Style changes over time

and

can be described over time.

Faras with its
superimposed frescoes (Michalowski 1967). Their sequence and helpful
inscriptions allow four successive styles to be identified between c. A.D. 700
and A.D. 1200. Labelled by the dominant colours used, they are the violet
(700-850), white (850-900), red and yellow (900-1000) and many-hued
(1000-1173 at Faras only!) styles. The major ateliers were always at Faras
although provincial ateliers following the dominant style of Faras, seat of the
bishop, have been identified. Each style evolved locally and characteristic
influences from the outside are known, from Coptic at first to Palestinian in the
white period, Byzantine in the red/yellow period, and intensified Byzantine in
the many-hued period. The problems of style, barring a few details, are solved

The most famous example in

stylistic seriation

Africa

may well be the Cathedral at

at Faras.

But this case is uncommon. Usually a series is established, it is known that

must have some antiquity, and then it is classified. The best known example
concerns the seriation of Benin copper-alloy works. Traditional data ascribe
their origin to a time well before the Portuguese appeared on the coast, and
brass heads continued to be made shortly before 1897, when the Benin palace
treasures were looted by the British expedition. The first stylistic classification
of von Luschan (1919) was replaced by one which W. Fagg and his school
elaborated from c. 1950 onwards (Elisofon and Fagg 1958:57-8, 646; Dark
1975), and which gained wide acceptance. In its mature form, five periods
were distinguished for the heads. The fourteenth to fifteenth centuries knew
very thin brasses of very high technical quality with high collars under the
chin. By the first half of the sixteenth century the collars were rolled; by the
late sixteenth and seventeenth centuries high collars reached over the chin.
The eighteenth century saw the addition of a base flange to the heads and the
metal work had become much thicker. By the late nineteenth century the cap
acquired wings. All through this evolution the faces became more and more
stereotyped and occupied a smaller and smaller portion of the whole. The
typical Benin plaques were believed to have been made between the 1550s and
1700. Dark divided them into three stages, according to background and
relief. The chronology was based in part on oral tradition, in part on passing
it

references by authors.

From the 1970s onwards the scheme has been challenged on various
grounds (Shaw 1978:172-84for a summary; Tunis 1981;Fraser 1980, 1981a).
The more naturalistic heads were assigned to the onset of the sequence because
traditionally it was claimed that the art came from Ife which had a naturalistic
tradition. But the Ife connection is evidently much less direct, if there actually
was one, than had been assumed. The sequences for plaques and their terminal
dates have also been questioned and even been turned completely around by
some scholars. Features such as flanges may have appeared side by side with
flangeless works for quite a while and early heads or rolled collar types might
be variants referring to certain situations or belonging in the beginning
perhaps to a rival workshop (there was one at Udo according to tradition). In
short, convincing proof of Fagg's scheme is lacking. Many more pieces must
95

ART HISTORY

IN AFRICA

be dated by thermoluminescence of the clay cores before the matter will be

settled. Today we know that in the thirteenth century tin bronze arm or leg
rings (manilla) were manufactured in Benin and that established
nineteenth-century works were in leaded brass. The heads are all brass. Some
supposedly sixteenth-century items have been dated to that century but items
that present particularly close similarities to Ife have not always turned out to
be of early dates, some perhaps even being from the seventeenth century.
Traditions tell us that one king ordered a stool to be fashioned as a replica of
one made two hundred years earlier. Thus revivals of older forms in plaques or
heads are not to be excluded either. The whole sequence will certainly be more

'-

^-^Eryxsti&s

--

Fig. 5.5 Senalion of Benin brass heads, as proposed by W. Fa^fi and P. Dark. The appearance of
elements, full lines. Progression from A to H spans perhaps four centuries

96

new

dislinctwe

STYLE

Plate 5.7 Head. Brans. Use unclear, perhaps on ancestral shrine. Renin, Nigeria. Rniish Museum. Height 21cm.
Supposedly Early Style' Hence 15th or 16th c. Rut the four marks and hairstyle indicate a foreigner. Hence this is a
trophy head. Two such heads in Renin are linked to kings living c. 1500 and just before 1800, respectively (cf. Ren Amos
'

19S0:IS). Date

unknown

complex, when unravelled,

if

it

ever

is,

than the logical sequence posited by

style seriation alone suggest.

Seriation can be applied to the other arts just as

sculpture. Dated series of the portals of congregational

it

is

to painting or

mosques in the Fatimid

97

ART HISTORY IN AFRICA

A Mahdiya. Tunuia 913; R al Aqmar. Cairo 1125; C Rathars, Cairo

a
126719. All portals for congregational mosques. The portal is flatter at al Aqmar hecause it had to be integrated
The conlinuay of
fagade on the street (the first in Cairo) and it is deeper at Baibar's mosque where a supported a minaret.
Fig. 5.6 Fatimid portals. All stone masonn.\

the basic design

98

is

remarkable

STYLE
period show the appearance of a portal first at Mahdiya in Tunisia c. 916, its
development in the al Hakim mosque c. 991 in Cairo, its incorporation in the
first true street fagade at al Aqmar in 1 125 (Cairo), and its final elaboration in
the sanctuary built by Baibars I, the first Mamluk ruler and dated to A.D. 1267.

The

original portal

may have been

inspired by a

Roman

triumphal arch. This

led to the sequence in Cairo, but also to the

composition of the minaret of the Banu Hammad at their Qal'a in Algeria. By
the onset of the Mamluk period the rulers began to build madrasa and tomb

structure at

Mahdiya not only

complexes rather than mosques. The portals of these new types of building are
strikingly different, as, for example, in Sultan Hasan's complex of 1356.
Apparently the model for these is the portal of the Gok Medrasa at Sivas
(Anatolia), a creation of the Seljuk Turks. But the notion of having a portal in
front of a religious complex was familiar in Cairo from the Fatimid
congregational mosques, even if the shape of the new portals was Anatolian.
It is significant that the Fatimid portal series was linked to congregational
mosques. The madrasa/ tomb complexes were very different in overall
conception as well as intended use and had come to Egypt from the Middle
East. Thus there was no replacement of one type of portal by another, but the
replacement of one type of building by another as the object of the ruler's
attention (Sourdel-Thomine and Spuler 1973: ill. 162, 163, 168, 176, 288, 295;
text:244, 249, 329).

The stylistic series for Fatimid portals is convincing, because the series is
dated, organically linked to one type of building and to the preferences of one
dynasty. Without dates the series might have been arranged quite differently,
Mahdiya might then have been seen as a provincial copy of Baibars' portal, for
instance, and been last in a series rather than first. Such possibilities of
arranging undated elements in other equally convincing series recalls the
puzzle from Benin and highlights the dangers of using the fact of a stylistic
series as a real historical chronology. A styUstic series always represents an
hypothesis of sequence over time. It is a logical device, much used by
archaeologists, for instance, in the seriation of pottery or types of tombs. The
simpler the shapes, and the denser the series, the greater the chances that the
sequence has some historical validity. But seriations should never be confused
with proven chronology. To the contrary, they should be tested by every

known

device in an effort to falsify or confirm the hypothesis.

SHAPE IN TIME
of style will always remain subjective, but the degree of objectivity
can be considerably increased by a clear exposition of an author's reasoning,
either by the use of a highly formal technique of analysis such as morphological
analysis or by a clear expose as to why some works are seen as prime works and
others as their replications. The reasoning can then be checked for circular
elements, low probability (as when a supposed replicate could well be the
original or vice versa), inconsistency and implications from the point of view of

The study

spatial distributions

and known

historical

developments

in the area studied. It

then be easier to see just where research stands on a given stylistic

problem. A typical example is the stylistic sequence sketched out for the
sculpture of a small coastal group, west of the Cross River, called the Eket.

will

99

ART HISTORY

IN AFRICA

Neyt (1979:24-6) studied one collection and distinguished between archaic

forms and later derivative forms. Archaic sculpture was characterized by such
features as the heart-shaped form of the face and often by ray-shaped
appendages around it. The classical form following showed a more definite
precision both in the cutting of volumes and in their assembly, while its
polychromy has a 'delicateness of tone and an equilibrium of colours' used as
slip and as highlights that is absent from later works. Later work, from the
present century, simply lacks the perfection of classical work, both in the
carving and in the painting. The rationale for the archaic end of the sequence
was that such archetypes were common to the Eket and all the neighbouring
groups sharing in the same general culture.
The main specialist for the Cross River area, Nicklin (1980), approved of
Neyt's stylistic approach, and so we have a consensus of two. Presumably he
understood just what 'delicateness of tone', 'equilibrium of colour', meant as
any reader could by comparing classical and late works. Nevertheless, more
specific statements about muting or not of colour and about colour oppositions
and complementarities would help.
How is one to go beyond such a consensus on a local sequence? In this
case, diligent study of a larger corpus and morphological analysis would render
the argument more specific. A detailed search into the historical conditions of
the Eket and their neighbours can be made from the late eighteenth century
onwards and should not make the proposed sequence improbable, but rather
strengthen it or bolster an alternative. In the end, this local sequence must be
placed in a much wider West African framework. Such problems are beyond
the stage of style seriation.

100

They

are discussed in Chapter 10.

CHAPTER

SIX

THE INTERPRETATION OF
ICONS
ICON

AND CONCEPT

study of medium, technique and style has still not considered the gist of an
Art objects are physical images which are the materialization of
mental images associated with definite meanings, i.e., icons. Whenever I want
to stress the link between the object and the visual concept underlying it, I will
use the term icon. Because of their meaning, art objects are understood,
decorated or handled in a special way. The Ardra board (Plate 1 1) is not just a
art object.

board:

it is

board for

ifa divination.

It

has an effigy of Legba or Eshu, the

god of Chance and Fate. It carries symbols of divination, and is used

Most of the images around its rims refer to meanings now lost to
us. We see only gesturing men and women, animals and a few objects such as
tusks - the latter representing wealth - but we cannot make sense of the whole.
The user in the seventeenth century could, and probably all these meanings are
trickster

for divination.

connected with Fate.

While the quality of image is evident in the case of the Ardra board, it is
not so self-evident in the case of household objects such as a wooden milkpot
from the Great Lakes area (Sieber 1980:ill. 277). Yet that, too, is an image that
stems from and leads to concrete visual concepts. When the people of this area
think about a milkpot, it always has a particular shape familiar to them and not
that, say, of the calabashes in which West African pastoralists store their milk.
The mental image of the object exists before the object itself, an idea that
guided the maker. In turn, the object made impresses itself as an icon on the
mind of the viewer. It is known that visual concepts are more powerful than
auditive stimuli because the memory codes them twice, once as a memory of
sound and once as a memory of image. They are more immediate than most
other types of memory because of their concreteness. Visual concepts of this
type also persist more over time, because the object created lasts and acquires
its own independent life, independent from any mind, unlike auditory stimuli

which are only preserved in the mind. Hence visual images, icons, can have
great impact on concept formation (Ohnuki Tierney 1981).

A Kuba man once described his unexpected encounter with a nature spirit
near the river in the early morning mists. He described it in detail. But the
account from which this episode stems goes on to explain that it was the
trickster, who had donned the costume and the hwoom mask that represent
nature spirits. To the Kuba, the idea of a nature spirit looks like a bwoom
dancer. The concept is visual and because it is, it becomes concrete.
No one doubts the existence of nature spirits, because people know
what they look like. This is no different from the popular conception of the
101

ART HISTORY

IN AFRICA

devil in the West. We know what he looks like, with his pitchfork, tail and
hooves, nimble as can be, matching black and red hues. The concept is real
because it is visual. Its concrete character saw to it that its reality was rarely
questioned. Yet there was a time in the first centuries A.D. when this stereotype
did not exist, and devils were portrayed as refulgent angels. Similarly, we
know what dragons look like, although there never was a dragon. So it is
natural that a nature spirit among the Kuba is perceived as given by
experience. A little further reflection shows many natural items to be seen, not

Gombrich and others have shown that we

had been portrayed. The visual arts therefore carry

as they are, but as they are portrayed.

perceive the world as

it

great weight in any culture.

The existing icon is a link in a chain of images running from the artist's
mental image, over the icon, to the visual image it inspires in the viewer and the

who later commission or execute just such a work as inspired by the

icon becomes the source for an artist's mental image. The whole
process is shared by the community of patrons and artists. The visual concept
is a 'collective representation'. In this sense, works of art are collective
products, cultural products, and this sense is very real. It is not therefore wise

artist alike,

icon.

The

medieval art in Europe is communal whilst

At the most, one can say that expectations of visual
concepts are less defined now than they were before in practically any art
anywhere at any time. Industriahzation has allowed us to make more and more
perfect, but banal, copies. Hence the dissociation between a unique art object
and some other thing which is mass produced.
Visual concepts such as the Ardra board or a bwoom mask are also
statements. Even though the tray's visual impact is immediate, it is composed
of motifs, associated with the divination. Chance (the circle) is there. Fate (the
top figure of Eshu/Legba), the World (carvings around the board). As a
statement, visual objects represent a theme or a subject. This sculpture
represents a hunter, that painting a nativity. The theme is made up of details,
to argue that art in Africa or

modern Western

art is not.

r-

THE INTERPRETATION OF ICONS

the motifs, which are visual units often isolated in space, as in this case, but not
necessarily so. Decorative borders as a whole outline space, but within them,
the motif often forms a pervasive repetitive pattern.
Isolated motifs clearly are the 'words' of the statement,

which is the icon.

no problem here, but patterned ones do. For they
must be the units seen by the artists and their public, not the units stemming
from the visual habits of the scholar. Sometimes there are obvious problems
here. The keyhole windows in monolithic Ethiopian churches make no
immediate sense to the Western eye, yet they represent the opening of an
arched window with pediment and corbelling. The Ethiopians did not 'see' the
arches, corbels, pillars and bases around the window, but they saw the space
delimited by them as a unit. To Western viewers, the effect is rather as if the
background had become the motif, and the real motifs the background. In the
rock-hewn churches, the shape of space was reproduced: that is, the keyhole
Isolated visual units pose

shape.

THEME AND MOTIF

effect of the parts yields the overall visual concept of the icon.
Often motifs and their compositional arrangement are all part of the
conventional composition rendering the theme, 'the' visual concept. Consider
the Central Angolan (Chokwe) statue of the ancestral hunter Chibinda Ilunga
in Plate 6. 1. He stands holding his gun in one hand, his staff of a wanderer in
the other. His hat tells us he is a chief, and the horns or figures on the hat show
his control of supernatural powers. The horns are filled with magical
substance, the figurines are his invisible spirit, familiars. His shoulder bag
identifies him as a hunter and not as a warrior. The size of his hands and feet
tells us of physical strength and endurance. The stare, the flaring nostrils, the
wide, thin mouth, and even the ears, tell us of power also to be feared. More of
this can be understood as the theme of Chibinda Ilunga becomes known. The
name refers to the supposed founder of the civilized world, a hunter appearing
from the East who united with the then-ruling woman to found a mighty
empire. Even if we know this, we still cannot infer from the statue alone what
the details mean, not even by comparison with the other Chokwe works. Local
informants have to tell us about the meaning of the facial features and hands.
All the motifs here are fused into a single composition, rendering a single
theme and they can be found with httle variance on other such statues. A
glance tells Chokwe they all represent Chibinda Ilunga (Bastin 1965). On the
other hand, there is absolutely nothing in any subject or theme that requires it
to be rendered in a particular way. The relationship between the concept and
the rendition is almost as arbitrary as that which exists between sound and the
meaning of a word. In the Lower Niger the hunter figure was represented as
striding in movement with a kill over his shoulders, in other sculptures he trails
his dog. But the variance can be much greater than this. We imagine that a
mask of a person should at least represent a face, but this is not necessarily so.
Just west of the Cross River, along the coast, among the group known as Eket,
one series of masks looks like a huge sandwich board decorated with gay
floral-like patterns (Neyt 1979). Yet the mask represents a person. It is
important to stress this point because the relative, arbitrary, character of a

The cumulative

103

ART HISTORY

IN AFRICA

Use unknown.
Plate 6.1 Chibinda Ilunga. Figure. Wood, cotton belt, natural hair, fibre, yellow bead (bard).
Chihmda Ilunga is the mythical hunter who married Rweej and thus founded the l.unda empire. North central .Angola
(Chokioe). Museum fur Volkerkunde, Berlin-Dahlem. Height 39cm. .Acquired 1879

104

THE INTERPRETATION OF ICONS

P\Me 6.2 Hunter figure. Copper alloy. Use unknown. Found ai Benin
IS

totally unlike

1897. Bniish

Museum. Height 21 cm. The

any other Renin work. Date of manufacture unknown, could he 15th

to

19th

style

c.

105

ART HISTORY IN AFRICA

means that when several representations of the same theme are
quite similar, they must have an historical connection.
As visual units, motifs can occur in other compositions than the one where
representation

originated. Women with their hands on their bellies and men with
arms occur on the Ardra board. But they are quite common in
nineteenth- and twentieth-century Yoruba sculpture of the same general area
as well. The royal hat of Chibinda Ilunga is the main element of one of the
Chokwe initiation masks and occurs on the upper back rung of thrones. Motifs
then are real visual units, and yet the term is also used, perhaps unfortunately,
in a vague sense. An arm is raised and one can talk about the 'raised arm' motif.
The motif is no longer self-contained. It may not have been a visual unit to the
artist, to be separated from the figure to which the arm belongs. It is a motif
within a motif The problem is not that motifs may be internally complex, but
that the term is used to designate compositional particularities which may not
have been visual units at all to the makers. Thus, a sentence referring to the
motif of repeated volumes is very different from one identifying a hat as a
motif. In the first sense, the term refers to an analysis of composition, in the
second to a demonstrably separate motif The confusion introduces us to the
problem of interpretation.

they

first

raised

INTERPRETATION
Meaning is never self evident, even though it often seems clear enough - this is
a mosque, that is a crucifixion, that again is the statue of an ancestor - but it is
clear only because it has become familiar and we have been told what these
buildings, paintings or sculptures meant. The icon has become part of our
culture. Apart from the culture, the meaning is as opaque as anthracite coal.
first step in dealing with interpretation, then, is to give an objective
description of the object, not refer to meaning. 'Kneeling woman with a bowl'
is an objective description of a kabila figure of Shaba (see Plate 5.5). 'Beggar
woman' is perhaps compelling but it is totally wrong (Maes 1938:78; Olbrechts

The

1959:71, 106-7). 'Monkey with bowl' is correct for a figure from the southern
Ivory Coast (Baule), 'Trained baboon begging' is wrong. Interpretation is
therefore often dependent on data not contained in the art object itself and
poses critical problems. Figurative representations pose especially formidable
problems of interpretation, unfortunately all too often overlooked.
The Saharan rock paintings (5000 B.C. and later) illustrate the question.

At first, scenes were labelled with imaginative names such as the 'tooth puller',
'martians' (a whole category of faceless representations), and even 'Josephine
sold by her sisters' The description accompanying this composition tells how a
woman was sold by her sisters to strangers while all we see are four figures! The
inspiration for this particular fancy seems to have been a mix between Joseph
.

in

Egypt and thoughts about African bridewealth (Lajoux 1977: 1 10-1 1). Later

scholars tended to be

more circumspect but

still

accepted the interpretation of

Plate 6.4 Rockpainting. Family scene in and near a shelter. Inside a child plays under a stand for a bowl. Other bowls,
child.
lie on the floor. A person rests in the shelter. In front, nearafire(?) a woman and
Below them another woman addresses a child in front of another shelter. Sefar, Algeria. Earlier than 2 000 R.C

gourds, pots hang from the wall or

Pastoral neolithic. The penodtzation

is

106

The styles of Plates 6.4 and 6.3 vary not as a succession of a

and space. Note the various perspectives used

very approximative.

single style over time, but as different styles in both time

THE INTERPRETATION OF ICONS

and us tethered calf. The cow needs

Plate 6.3 Rockpainimg. Family fcene: tnan. woman, two children sucking, a cow,
m order to give milk. Karkur Talk, Jehel Uwainat, Libya. Pastoral neolithic. Earlier than 2 500 B.C.

to see the calf

Compare

icilh

PldU

fi

-/

107

ART HISTORY

IN AFRICA

from an initiation myth, because they were so

understood by a Fulani scholar, who thought he recognized episodes of the
Fulani initiation myth in them. But so many centuries separate the Fulani of c.
A.D. 1900 from the scenes dating earlier than 2000 B.C. that the one cannot
shed light on the other (Lajoux 1977:106). We must therefore restrain our
labels and commentary to unmistakable objective elements: 'a hut with a man,
a child, a bowl, a calabash and another object'. It means that historians will
learn less from Saharan paintings than they once hoped for, but at least will not
mislead others.
In some cases the interpretation was linked to much wider issues. In the
Brandberg area of Namibia exists a famous rock painting, labelled 'the White
Lady from Brandberg' by the inventor Maack and later Abbe Breuil, a famous
French prehistorian (Breuil 1955; Jacobson 1980). Breuil had seen
reproductions of it in 1929 and 1942. In 1947, Field Marshal Smuts helped him
to visit the site. On his return, he gave a lecture at Windhoek* and 'confirmed'
that the 'white lady' was of a European or Mediterranean type, perhaps Cretan.
That he had already seen from the photographs and told to Smuts before going
to the site. He now concluded:
certain scenes as episodes

Allied with the grey and extraordinary figures in the Brandberg

number of figures of strangers

of different races, painted in Southern Rhodesia and the South East of the

arises the question of the origin of a certain

Union.
B.C.

The trade winds: Monsoons, Sofala civilization of Ur. Sumer 3000

The Fort Victoria frescoes. Chibi Impey [sicj cave certainly of the

same

style as the

Brandberg and point

to influence

coming from

that

region.

The

spiritual landscape of Africa.

A glance at any illustration of the figure shows it to be male and black (Willett
1971). Its datation by amino acid counting makes it 1 200 to 1 800 years old nowhere near 3000 B.C. - and this is no surprise, as Breuil mentioned shards of
pottery on the site (Iskander 1980:218). There is no close relation to
Zimbabwean rock paintings at all. The illustrious archaeologist was so blinded
by racial prejudices that he apparently could not even see what the
representation showed. His error was demonstrably monumental. But was it
that

much

greater than those of the scholars

paintings, or of

all

those

who

who

labelled the Saharan

attribute awe, fear, mystery

and such

like

emotions to African sculpture?

The difficulties of interpretation cannot be over-stressed. In Old

Egyptian paintings a typical composition shows a larger sized person
surrounded by smaller people occupied with activities of daily life; these are all
tomb

scenes (see Plate 6.5). Is the larger figure the deceased supervising work
on an estate of sorts, or is the scene merely recording the theme of a daily round
of life and divorced from the larger figure? We often do not know (Gombrich

* Breuil papers f A461. Notes by Abbe Breuil for his lecture, Windhoek 1947. From the
Department of Historical Papers. University of the Witwatersrand Library, Johannesburg.
thank Dr H. Scheub for the text.

108

THE INTERPRETATION OF ICONS

reJltlUt>TfiTiiiPff'tl<<l^*

!g

Plate 6.5 Tomb of Menna. Shaikh 'Ahd al Qumah, Thebes. Scenes from farm life are shown with figures of Menna
under a canopy. Upper right two registers shoiv managers from his estates arriving with produce required (not shown on
this

and

photograph). The upper group

is

honoured because

their remittances are satisfactory.

Those of the lower group are not

managers being servants or

they are being beaten or will be. Different authors explain these scenes differently, the

is a servant praising Menna. The

The Scribe of the fields of the Master of the two lands, North and South: Menna'. It
IS unclear whether all the scenes must be read together or not. Is Menna supervising the arrival of the managers and the
threshing? Is the person carrying vessels bringing them to Menna or putting them on the side of the field? Is Menna shown
in his role as an overseer, actually on the estate? Or is his canopied figure something to be seen completely apart, a
representation of the deceased? Tomb no. 69, southeast wall, detail ofwallpainting. XVIIIth dynasty, c. 1420 B.C.

guests

and

inscription

the person pleading pleads for himself for the person being beaten or

m the top canopy reads:

'

I960: 122-3). Frescoes in churches such as at Faras do not depict scenes from

we designate them as such. They depict highlights

The location in the church and the choice of episodes

the hfe of Christ, although

of the liturgical calendar.

proves it. If we relate the motifs of the Ardra board to Chance, Fate and the
World, we do it because such boards are still made (see Plate 2.3). Fieldwork
has shown them to be part of the tools of the ifa diviner and to have such

meanings, but as we saw, we cannot thereby still interpret all the details of the
Ardra board; we might be anachronistic. If we interpret ivory as wealth it is
because we know that ivory was exported from there at that time as a valued
product of trade. Our interpretation of the whole board, however, will
probably always remain fragmentary.
How does an outsider reach a valid interpretation? The ideal would be to
know the total sum of local interpretations, which may vary from person to
person, but revolve around a common intellectual and emotional core, the
'collective representation'. No ethnographic report is that detailed. At best we
have hermeneutic exegesis by an insider. Thus, among the Dogon people of
Mali, Ogotemelli, a blind sage, thought and talked a great deal about his
109

ART HISTORY

IN AFRICA

culture and interpreted myths, themes, motifs for M. Griaule (1948) and his
colleagues. In turn, they expanded these and checked them with other

inhabitants of Ogotemelli's settlement.

The interpretation of Dogon sculpture

more authentic than outside guesses but it is a personal statement

which required careful checking with other Dogon. Ogotemelli's views
certainly cannot be extrapolated to interpret other works of art in other areas as
has sometimes been done. They do not allow in the name of symbolic
equivalence, for instance, to equate 144 holes on a vessel with 87 or so, as some
certainly

is

The visual evidence remains the most important!

Often we do not have much ethnographic information, or it deals only
with usage and not with a detailed reading of the iconic statement. Moreover,
for many older objects, we can only speculate by analogy, with all the dangers
inherent in that procedure. Here there exists in fact a deep difference between
the literate cultures of the oikoumene and the regional traditions. Christianity,
Islam, classical antiquity, Pharaonic Egypt, have left us a treasure house of
texts by which we can venture to interpret icons, if they do not, in detail. This
is especially true for all religious works. Elsewhere the researcher will always
be faced with difficult choices of interpretation and it behoves him to make the
grounds of the interpretation quite clear.
researchers have done.

ART LANGUAGE
make

Icons

all

sorts

of statements.

They can be

narrative,

signs

of

comments. But all icons need not make statements. Some

motifs do not. Moreover, icons always evoke emotional appeal. In narrative
compositions the motif becomes an element in telling a story. The Faras

identification,

of the shepherds aroused by angels, the holy

of the kings. A few brass plaques of Benin
represent historical scenes such as 'return of the Igala war' and their

painting of the Nativity

family

may

at

tells

the manger, and the

juxtaposition

may

visit

well have evoked narratives,

now completely

lost

(Ben

Amos 1980a:23, ill. 21). Sometimes narrative becomes mere anecdote as in the
Kuba cup portraying a man holding a cup which relates to an individual who
was known for his propensity to imbibe. On the whole, narratives were rare
outside Christian art. Compositions were not very frequent on a larger scale
and usually the representation of a single person or animal was to carry the

whole message.
Statements indicated by signs, usually signs of identification, were by far
common. Thus, attributes of Christian saints identify them: St
Andrew carried his cross and had tousled hair (Wessel 1965:166-7).
Differences in hairstyle, beards, sideburns identify different saints in Coptic
painting. The attributes of the Chokwe figure of Chibinda Ilunga identify him
as such by his gun, bag, staff, hat, figurines or horn, hands and feet. In toto,
they are a statement explaining who Chibinda Ilunga was and what the concept
the most

is similar to the statements made by Chokwe chairs where

every rung may be carved. Scenes of initiation, fights over women, daily
preparation of food, trade, caravan life and sex are common. Each of these
groups of icons seems to tell a separate anecdote, but the scenes can all be
related to the concept of the chief as the upholder of ordered social life
(Kauenhoven Janzen 1981). The great difficulty in interpreting such icons is

represents. This

110

THE INTERPRETATION OF ICONS

the complexity of the statements
enough, but the relationship among

made. Each scene

them

is

deciphered easily

not evident. Usually the scholar's

tendency is to link all the details to the statement of the theme, but that need
not be the case. The interpretation tends to remain hypothetical as long as
actual informants living at the time and near the place of fabrication are not
is

available. Yet stronger grounds for linking parts to the whole can be obtained
by comparison. For instance, if different Chokwe chairs show the same limited
number of subjects - as they do - the chances that all are related to the theme of
kingship are higher. At the level of cultural analysis a link of all these
statements to the central theme of kingship is valid. Yet it may not be valid at
the level of the intentions of each artist, his intended composition of a
particular chair. A scene invented by one carver, appreciated by others and by
the public, could become a favourite theme in later chairs even though its
connection with the concept 'kingship' was rather tangential to begin with.
The accumulation of scenes may not reflect kingship, but only the special
character of a throne as an object of prestige or merely the mastery of the

carver. Eventually the very existence of the chairs would lead to a visual
association of the scenes they portrayed with the notion of kingship although it is

scenes on the rungs were not intended from the beginning to

be portions of a detailed and purposeful statement about the nature of
kingship.
Motifs as signs often led to identifications that became portraits. A tattoo,
in
a style of hair, a horn in hand, a piece of jewellery or clothing, placed a statue
likely that

many

group and a type of social role as clearly as any attribute identified a

Coptic saint. Beyond these general identifications, portraits of individuals
could be meant. They were never made from nature, but they carried signs
which identified them as such for the cognoscenti. A Kuba (ngongo) statue of a
pregnant woman was the portrait of the daughter of the carver who made it and
who died in childbirth (see Plate 6.6). The sign of pregnancy was the only
indication of this visible to outsiders. The sculptor himself claimed to have
rendered the features of the face, conventional though it looked to others.
Kuba royal statues refer to individuals by impersonal signs. The founder king
a social

identified by his rows of emblems, which later kings lacked in quite such
profusion (see Plate 8.4). But essentially it was the emblem placed in front of
each statue that identified it, rather like a label. A game board, an anvil, a
woman, drums with different decorative motifs for different kings, all

was

identified the statue. Physical differences barely appeared, one only in the set
shows more than the ordinary obesity (an ideal of Kuba kings) by a slight

indication of rolls of fat in the neck. But in general the physical representation
merely refers to an ideal, just as rolls of fat in the neck of masks of the Sierra

Leonian Sande Society refer to the

ideal of feminine

beauty and well-being

Portraits thus existed but not

mouths denote wisdom.

Perhaps the representations of Ife heads may have been
of northern
lifelike. If so, they were the great exception. This is also true
African arts in general where identification occurred by general signs and by
inscription of a name. Only Roman portraits and the painted effigies of
mummies in the first centuries A.D. were really intended to be lifelike.
The use of icons as incidental comments detached from the main

there, while their small

lifelike portraits.

statement, or the main theme, was not

unknown.

A Coptic

painting of David

111

ART HISTORY IN AFRICA

Plate 6.6 Figure of a pregnant

village in

Ngongo andMheengi

Unlike any other

style

woman. Wood. Used as a memento. Keenge, northeastern Kuba kingdom (Ngeende

June 1956. Portrait of the daughter of the carver who died in childbirth.

area). Photo:

the area, this piece

is

unique

Persian
and Goliath shows Gohath, the bad man with an Arab beard and clad in
pamtmg
Ethiopian
the
is
armour (du Bourguet 1967: 170). A strident comment
a scene
of the prophet Muhammad, bound to a horse and led away by Satan,
Sellassie near
Berhan
Debra
at
themes
usual
the
between
sandwiched in

Gondar (Leroy 1967:35).

South of the Sahara, such statements are known from the performmg arts,
but have not been reported in the visual arts, probably because the allusion
remained esoteric and not because it did not occur.

THE COMPLEXITY OF MEANING

meaning of icons can be quite complex because of the situation in
which they appear. The well-known juxtaposition of proverbs with depictions
on goldweights in Ghana or potlids in Cabinda typify this (Appiah 1979;
use.
Cornet 1980). The concrete meaning of the icon was apparent only in its
To send a goldweight of the two friends shaking hands to one chief in one
circumstance meant one thing; to send it to another at another time meant

The

112

full

THE INTERPRETATION OF ICONS

woman

to her husband in one circumstance could

use in another circumstance. To decipher
interpretations it is therefore not enough to find the main field of meaning, but
it also becomes necessary to review the circumstances of use.

another.

potlid sent

by a

carry a message different from

its

Juxtaposing several icons produces even more complex statements in

which each icon acts as if it were a motif of the larger ensemble. The wall Kuba
built on the occasion of boys' initiations illustrates this occurrence (Fig. 6.2).
On it many icons were displayed together and the whole wall formed one unit.
Such a wall consisted really of a scaffolding of poles from different kinds of
wood, all symbolically significant, covered with raffia fibres. The wall was

shaped like a screen with three triangular peaks, well over nine metres in height,
the whole being sixty metres long and more. In front of the wall near its centre a
round armature covered with raffia fibre encircled an almost life-size female
figure and, in 1953, when I saw such a wall, two small figures were planted just
outside the enclosure surrounding the woman. In brief, the wall represented
the journey the boys to be initiated were to undergo once they passed it. It all
represented the mythical journey of Woot, the Adam of the Kuba. The
triangles represented hills; the mask on top of the central hill represented
Woot; the female statue was his sister, the primordial woman; her seclusion
was the icon for the prohibition of incest, while the tears on her cheeks told of
the grief of woman, forever separated from her kin. Masks on top of the right
and left hills were, taken in isolation, those of the king and of the nature spirit,
taken in conjunction they represented the king and the common man. At the
capital, that is precisely what they meant in mimes. Here they also stood for the
headmen of the right and left sides of the village, the village itself being a
symbol for society. When all three masks were considered together, the central

mask was the king, the right hand mask was the aristocrat, and the left hand
one the commoner. More than a dozen other carved objects representing
persons, birds, animals and a palm tree climber all had multiple meanings, one
in isolation and one in conjunction. Thus, some birds in a tree were the icon of

Kuha inuuition wall, Mapey, Zaire. Hetghi c. 9m. /953. On the wall display of masks, pole heads and other
some m composition (e.g. on and near the palmtree). 1 n front female fifcure (see PL 11.3) and two charms, ishak
ndweemy,/Mnc/ion(i//v similar to PI. 11.4, hut in very different style. The wall was a symbolic discourse, down to
details such as the three 'hills' and the different woods used in the scaffolding. It was a major teaching device during the
Fig. 6.2

icons

boys' initiation

113

ART HISTORY IN AFRICA

moon taken separately, but jointly here represented the boys to

be initiated (Vansina 1955).
Among the various purposes of this wall, the main one was teaching
symbolic reasoning to the boys. At the same time, the children's perception of
the canons of Kuba art were reinforced by such an exhibition at a time of
heightened emotions in the context of their passage from childhood to

the rain and the

manhood. Furthermore, the icons expressed complex relationships and

abstract concepts in a concrete way at a glance, thereby instilling visual images
which had much more lasting impact than a verbal explanation only would
have had, for reasons explained at the beginning of this chapter.
This thumbnail sketch must suffice to illustrate the complex nature of
meaning in art. It is almost always wrong to attribute just a single meaning to
an icon, as if it were a sign, like a giant letter standing always for the same
sound or better, as a hieroglyph. The meaning changed with circumstance and
equally important, the emotional impact of the concept changed with the
situation in which the icon was displayed, the number of times a person saw it,
and idiosyncratically as well. The icon was perceived as a unit and no one
differentiated

between

its

aesthetic, emotional

and conceptual appeals. In

this

icons played a role very similar to ritual. Communities all participated in the
same ritual but, in the absence of the dogma put down in a holy writ, all were
free to experience it differently. Hence the full meaning of an icon should not

be confused with the intention of its maker only. In Niger Delta (Ijo) art,
masks such as the familiar hippopotamus or shark masks were carved as
representations of the spirits, but were not to be seen by the viewers. As they
were worn facing the sky, viewers rarely saw them in full face. When studying
such masks, it is very useful to know this: they were carved as concepts, not as
objects to be seen (Horton 1965a). But this does not suffice. The context of
apparition, the iconic attribution of one mask in relation to other icons, the
range of variation allowed in the carving of such masks, and the range of
experience evoked in the beholders, are all equally important in assessing the
full

meaning of such

objects.

primary interpretation of representations is already so difficult to

establish, it is even more difficult to acquire valid information on the full range
of meaning in varying circumstances, and these facets of arts in Africa have so
If the

been studied in depth. The situation is better in hterate

environments, but even for major features of the monuments of Islam it is hard
to explain the overwhelming appeal of some forms such as domes (qubbat); the
sickle of the half moon (hilal) or the appeal of chandeliers in mosques, an
appeal quite equal to that of church bells for Christians. Bells are an
far

rarely

abomination to Muslims; only the human voice can call the faithful to prayer,
and so, when they captured the bells of Gibraltar, the people of Fes turned
them into chandeliers for their Qarawiyyin mosque (Terrasse 1968).

DECORATION: ART WITHOUT A STATEMENT

Not all motifs make statements. A famous example

is

the mappula or

handkerchief of Ethiopian madonnas of the seventeenth and eighteenth

centuries. The handkerchief in the hand of the virgin meant nothing in
particular. The motif had come to Ethiopia when reproductions of the famous

114

THE INTERPRETATION OF ICONS

Plate 6.7 Madonna Hode^iithna. By Anorewos? Pamiinson cloth fixed on wood (Hagenia abyssinica). Central panel
of an uon. Northern Ethiopia. Mss. 8l-U)IJi64 Museum fur Vdlkerkunde, Munich. Height 36-5cm; ividth 21 ^cm.
(iondarfml period, 1700-1750. The handkerchief (mappu\a) and the folds of the headcloth correspond to the ultimate
model, the madonna of Santa Maria Maf^gwre, Rome. Unlike others in this style no star occurs on the veil or on the
shoulder.

The madonna guarded hy two angels with

apostles at her feet,

is

a standard composition

(cf.

S. Chojnacki

1977:44-7, 56-6/j

115

ART HISTORY

IN AFRICA

Byzantine icon at Santa Maria Maggiore in Rome, so highly venerated by

CathoUcs at the time, were brought to the country. Ethiopian artists copied it,
gradually rendered it into a local setting and introduced local motifs. But the
mappula remained. It became smaller and smaller, but the madonna could no
longer be visualized without it, so it stayed (Chojnacki 1977:46).
Some decorative art may have developed in similar ways. Decorative art,
usually, but not always non-representational, has no meaning. It is there for
design reasons only, it enhances. Decoration was used to provide a frame to
enclose a space, or to separate spaces as in architecture. Decoration could also
organize space, either by covering it with a patterned web, giving repetitive or
geometric meaning to space, or by using isolated motifs as focal points

empty space, to suggest space as a

presence rather than as an absence. Decorative art is universal. All African
cultures used it from the rim of potteries, to the most varied surfaces such as
shields, bags, baskets, bodies, walls, pediments, ceilings. Decorative motifs
were made in all media by any means that could produce a line, a dot, a point. It
reached its summit in Africa as arabesques or sequences of geometrical figures,
with interlace (rinceaux) originally stylized leaves, stems and fruits and arable
script. These were used in isolation or in combination and allowed an aesthetic

regularly distributed over otherwise

expression of form as refmed as any other and as linked to technical and

stylistic skill. Unlike figurative art, arabesques, writing apart, had no specific
meaning, and their aesthetic was completely different, acting only as linear,
textured or spatial design. But it was no less a great art. North African
decorative patterns profoundly influenced some decorative patterns of West
Africa, and all those of the coasts of eastern Africa.
Decorative art was meant to embellish or to convey emotion, not to
express statements. Even if patterns were named, the names formed no
statements. The Kuba have some two hundred named patterns but the names

between the shape of the design and some natural shape such
an iguana or refer to the name of the inventor. A very few
indicate that perhaps once the pattern had a meaning. Thus one is called 'the
house of Woot'.
refer to analogies

as the track of

Fig. 6.3 Kufic script as omameni. Plaster on the wall of Sidibel

'In the

116

name of God'. 1296. After G. Mar^ais 1954:

fig.

Hasan mosque, Tlemcen,

221

Algeria. Iireads: Bismillah:

THE INTERPRETATION OF ICONS

Sometimes motifs could be used, not for their decorative effect, but for
name. A Kuba motif of an insect called 'god' was found under a rendition
of a crucifixion where it obviously was to be read as 'God'. This rebus-like use
was unique there. Several authors have claimed, however, that decorative
patterns isolated in space and juxtaposed, really formed a rebus or a script, but
the most detailed claims for both goldweights and Bini patterns, mainly on
tusks, have not been substantiated (Hau 1959, 1964; Dark 1981; Thompson
and Cornet 1981). Still, Egyptian hieroglyphs developed in this fashion and in
a reversal Arabic, especially Kufic, script became a major element in the
arabesques. The reversal was not total, for Arabic script kept its meaning and
their

such decorations are in fact inscriptions.

We have seen that on occasions a motif, even a figurative one such as the
handkerchief, can have lost its original meaning, and that decorative patterns
usually contain no statements, and yet there are isolated cases where a pattern
functioned as a sign and is a statement. Like many other peoples in Africa, the
Kuba use a free-standing pattern of guilloche (interlace) in double knot form
on a square, called the imbol. When it is used on the back of a certain type of
mask, however, it indicates a relationship to the king. In other contexts it has

meaning. Free-standing designs on Masai shields, some tattoos in

and Mamluk crests indicated group affiliations. Full decorative
indicated position and rank as in the Akan states or in
cloth
patterns on
Dahomey. Such occurrences should warn the student not to dismiss decorative
art too readily without further enquiry.

no

specific

central Africa,

THEME AND MOTIF

IN HISTORY

motifs are the product of historical change. They often develop

and vary independently from each other, while constellations of motifs will
sometimes stay together for amazing lengths of time and spread over great
distances. As for themes, they live as images and as tales. An instructive case is

Themes and

the Coptic icon representing an equestrian figure slaying a crocodile (Plate

The figure in imperial garb was supposed to be Horus, the crocodile

6.8).

Sebk. As an icon it belongs to a series of martial equestrian figures that

developed in Sassanid Iran and in Egypt itself Later, with Christianity, Horus
became St George and Sebk the dragon or a snake. The icon became the theme
for the legend of St George saving the maiden and both legend and image are
found far and wide over Christendom, from Russia to Ireland, from Egypt to
Ethiopia and Scandinavia. But the tale spread further even without the icon.
The same tale explains the foundation of Daura, the oldest Hausa city-state in
northern Nigeria, the first foundation of the empire of Ghana as told about
Wagadu in Mali, the beginning of kingship in Taqali in the Nuba mountains of
the Sudan (De Grunne 1980:27-34; J. Ewald, personal communication). The
tale can be found across the Sahel and the Sahara from the Nile to the Atlantic.
It spread to Europe with the associated icon, and without it south of the
Sahara.

The history of the fish-legged figure in southwestern Nigeria (Yoruba and

Benin) tells the story of a complex motif (Fig. 6.4). In Benin it appeared on
over a hundred objects in ivory or brass, on bells, on wearing apparel, on
bronze plaques and in Yorubaland almost as great a range of settings for this
117

ART HISTORY

IN AFRICA

Plate 6.8 Horus and Sebk. Sandstone. Coptic. Miisee du Louvre. Height 49cm; length 32cm; width 7-8cm. Fourth
century. Horus in the pose of a Roman emperor spears the crocodile, an allegory for Good triumphing over Evil. This
forerunner of the iconography of St George

is

motif can be found. It was already to be seen on Yoruba ivories that reached
in the seventeenth century. In Benin, the central position of the figure
indicates that it represents either a king or a god. It wears crossed baldrics,
crisscross straps - signs of high rank - and its skirt ends in undulations,
rendered by the Yoruba as scallops. Some Bini take it that the figure represents
their Neptune, Olokun, others take it to be an old king who became paralyzed
in connection with Olokun. In any case, the interpretation is uncertain and
shows all the signs of being speculation about the figure rather than an
inspiration behind the realization of this motif. In Yorubaland as well,

Europe

interpretations

seem

speculative.

Fish-legged figures, sirens for instance, were very

common

in

both

European and West African art, but fish-legged figures grasping their own legs
with their hands were rarer. Only the Yoruba have this variant, Benin does not
118

THE INTERPRETATION OF ICONS

Fig.

6A Fish-legged person grasping its own legs.

Museum. Dale unknown. Here

Mediterranean. The conclusion

Ivory armband (detail), western Nigerian, perhaps from

Owo.

British

The whole motif derives from the Hellenistu eastern

unusual subject and a variety of elements remaining stable in the

the legs are not fish but reptilian.

is

valid because of the

composition: (a) self-graspmg fish or reptilian legs; (b) baldnc; (c) deep navel; (d) scalloped skirt.

On

other pieces the

complex includes further waistbands and segmented necklaces. The motif was common in southern Nigeria generally and
is attested before 1674. Such similarities cannot occur by chance. This motif as well as five others link southern Nigeria to
the Hellenistu: world

know

D. Fraser (1972, 1981b), who studied this case, showed that such a
was known in medieval Europe from the twelfth century onwards,
where it was called Melusine, or, more recently, Neptune. By 1674, the motif
was in use on ivories in western Nigeria. A link with Renaissance motifs is
possible, but the European legs do not terminate in fish heads, nor are there
baldrics or deep navels, as in the African examples. The full similarities occur,
however, in the art of the eastern Roman empire from 100 B.C. toe. A.D. 300. A
figure from Begram in Afghanistan was shown to be the most similar. The
prototypes were probably current in the heartlands of the eastern Empire and
radiated by trade to Afghanistan, India and ultimately perhaps Nigeria. It is
not a wild thought since Coptic lamp imitations of not much later date have
been found in northern Ghana. In fact, alternative explanations do not really
account for the similarities. Independent inventions did not invent fish-legged
figures, grasping their legs, with baldrics, deep navels, scalloped skirts and
even segmented necklaces! The motif travelled with all of these features, over
these great distances and kept them over such astonishing lengths of time. The
case is especially convincing on two counts. The icon is even more arbitrary
than is usual - who ever saw a fish-legged person grasping his own legs? Parts
of the icon were quite common in the eastern Mediterranean, including Egypt
at that time. Baldrics, scalloped skirts and deep navels were normal
it.

figure

119

ART HISTORY

IN AFRICA

iconographic

attire

for

high military

officials,

even for the emperor.

theme in that area since hoary antiquity, and the link

between persons and fish (at least dolphins) was quite common at the time.
Self-grasping was a

fish-legged hero grasping his legs could have arisen easily in this milieu.
have seen that icons represent visual concepts that are often
statements involving complex meanings, statements such as those that are

The

We

revealed in poems, with their emotional appeal and layers of meaning, rather
than those that are common in prose. Cases show that meanings, themes, and
motifs associated in any icon at a given time are unstable. They can change
independently of each other and have their own histories. Historians should

never reduce meaning to a gloss nor consider a composition of motifs, a theme,

as a once-and-for-all creation. Such an association lies deeply embedded in the
whole culture that produced the work of art and the pursuit of the full
ramifications by which the composition came into being becomes cultural
history. For culture is not, as society is, the context of art. Art is an integral
part of culture

120

and can only be

fully

understood as such.

CHAPTER SEVEN

CULTURE AND ART

not the context of art, nor a matrix for art. Art is an integral part of
it expresses and thereby communicates to others. Visual art
works are the visual realization of visual concepts held in a community and
realized by the artist. The concepts are visual, just as visual memory exists.
The icon then feeds back into the community's visual concept. This explains
why the seventeenth-century Ardra board is so similar to Ifa boards from the

Culture

is

the culture which

same area, even after 1900 (see Plates 1 1 and 2.3). The visual concept was alive
over all that time and it was collective.
Culture is the sum total of the ideas, aspirations, values and beliefs in the
mind of the people living in a given community (Spradley and McCurdy
.

1975:3-41). It includes not only ideas or even beliefs but emotional stance as
Most of this can be rendered by the expression 'collective representation'
and people's patterned reactions to them. For most of what is in people's minds
is held in common by them, just as visual concepts are, and just as language is.
There is nothing surprising in this, as the furniture of the mind is gradually put

well.

in place even from birth onwards. Babies not only acquire language, but
grammars and worldviews as well. They not only learn to see, but they learn to
see what is significant and to see in patterned ways. They not only learn to
remember, but to remember by a master code common to the community.

Arts are called expressive culture because they express the world of the
or through the creation of objects. Like language,
but unlike most speech, their statements
communication,
of
they are a mode
involve an expression in terms of forms linked to metaphor. Among the arts,
there are those which, like language, are statements over time: the performing
arts such as dance, music, oral art. And there are arts that once created are

mind either by performance

statements

The

one moment of time. These are the visual and tactile arts.
and culture is discussed here in several
the relationship of one art to the others, then the relationships to

at

relationship between art

steps. First,

the whole of the culture.

group

in

which

it is

The

examined

specific relationship of culture to the

falls

human

beyond the scope of this study. The reader

however, that culture is a relative concept because it presupposes the

recognition of a given human community. Since on the one hand every mind
contains some idiosyncracies, every person's culture can be examined. But
since a portion of the commonly held representations can extend over many
will note,

societies, culture can be applied to very large groups of people as well. Thus an
oikoumenical tradition where everyone accepts the basic tenets of Christianity
or Islam can extend over large portions of the globe, even while such tenets are
perceived and certainly expressed in very local terms.

121

ART HISTORY

IN AFRICA

THE VISUAL ARTS COMPARED

compared art objects in order to set up
Here we examine the mutual
stylistic series
relationships among the visual arts. Architecture, sculpture and the
two-dimensional arts are differently related in different cultures. Thus,
In previous chapters

and

we have

already

to discuss style in general.

architecture occupied pride of place in

all

of northern Africa; in Islam as in

Christianity, as in Pharaonic Egypt, as in pagan Rome, as in Ethiopia,

monuments came first. Sculpture, furniture and painting complemented

were used together. Muslim

of capitals, conchs, friezes. In
forms
sculpture included the non-representative
Ethiopia, most sculpture consisted of reliefs in churches and palaces.
Similarly, non-representative painting in Islam was used only to decorate. In
Ethiopia or in the Christian art of Nubia, as in Pharaonic Egypt, wall painting
was the most common form and was strictly integrated in the architecture, for
instance, particular scenes from the liturgical year had to be painted on

architecture

whenever these

different

arts

predetermined portions of wall within the churches.

This contrasts markedly with sub-Saharan Africa. In most regional arts
there was less of an organic hierarchy. One cannot really speak of the
dominance of any art, even though the literature often gives the impression
that sculpture dominated in West or Central Africa. This neglects
non-representational, two-dimensional art and wall or rock painting. In most
cases, sculpture

was not

as subservient to architecture as in the oikoumene,

and

few instances one can see how the conception of architecture, i.e. a set of
masses and interior space, was translated into a set of volumes. Figure 7.1,
in a

Fig. 7.1

Tomh

ofSidi Aissa. Meltka

Sculptural handling of architecture

111

le

Haul, Sahara. Algeria. Dried mud. Date unknown. Compare with Fig. 5.3.

CULTURE AND ART

at Melika in southern Algeria, strikingly illustrates how a
northern Sahara tomb building became a tomb sculpture. Studies of mosques
to
in West Africa (Prussin 1968) document the tendency for such buildings
while
Sahel,
the
parts
of
southern
become smaller and more shrine-like in the

showing a tomb

the latent sculptural qualities of

mud

architecture were

more and more

exploited.

Public buildings in

Gabon and Congo, such

as the Bwiti temples of

southern Gabon (Tsogo) or the men's houses near the Sangha River, were
simple constructions. Only the fact that they housed the carved posts and
figures really distinguished them from other dwellings (Bruel 1910; GoUnhofer
and Sillans 1963). In some regional traditions, however, architecture did come
into its own. In part of the Cameroons grasslands (Bamileke), for instance, the
sculpture of the posts of palaces was subordinate to the total architectural effect
(Lecocq 1953). Sculpture in ancient Zimbabwe also seems to have been partly
subservient to architecture insofar as it was destined to decorate the top of the
walls. Sculptures as posts in courtyards could also be found along the West
African coast, from Dahomey to Yorubaland, and to Benin, even though most
sculpture here was certainly not tied to architecture.
Painting, especially non-representational painting in West and Central
a
Africa, remained more important than is usually believed. It can be seen as
subordinate technique to decorate the walls of houses, panels, posts, or to be
the polychrome decoration of sculpture. But it came fully unto its own in
personal art and on textiles. Painting, like architecture or body art, is only
beginning to be studied, and it would be rash at this stage to claim that it was

unimportant.

abundantly clear for West and Central Africa that decorative,

schematic design was highly developed in all media as for example on textiles
or on ceramics. Just as in the European tradition, such arts were judged to be of
minor value to Africans south of the Sahara, and have therefore been less
It is

also

studied, but they can be crucial in some regional arts. Among the Kuba, for
instance, design is the essence of artistic activity; there is no Kuba term for
'art', but there is one for 'design' (hwiin), and it seems crucial to their
aesthetics. Decorative design was the most discussed, the most practised, the

most developed with regard to the solution of formal problems (Crowe 1971) of
fully
all Kuba art forms. Even though architecture and sculpture were
developed, was decorative art the dominant art? It was often integrated as part
of buildings or sculptures. The notion of a dominant visual art simply does not
apply to such situations where a hierarchy of either values or commissions has
not been developed, unlike the situation in the oikoumene.
The relationships between all the visual arts have often not been explored

Thus body decoration, a

West and Central Africa as

properly.

but in

spectacular art not only in eastern Africa

been related to textile arts,

well, has not

masking or painting, which seem to be quite closely related (Paulme and

Brosse 1956). In part, this is due to the fact that in Europe, this was never, until
very recently, recognized as an art. To a larger extent, such situations have
been influenced, often implicitly, by European notions of major and minor
such
arts. Those concepts have been transferred from Europe to Africa. But
distinctions are entirely ethnocentric. Calligraphy, a minor art in Europe, is
highly valued as a major art form in the Muslim world for instance, whereas in
123

ART HISTORY

IN AFRICA

the case of figurative painting, the reverse is true for most periods of Islamic
art. Such ethnocentric distinctions between major and minor arts merely
express the hierarchy of art forms in a particular culture at a given time. The

concepts are value judgements and useless for a comparative perspective.

Whether major or minor arts exist at a given time in a given place and what they
eventually are, are matters for investigation into the convictions of the
community that produces these arts. Meanwhile, a study of the relationships of
the visual arts in given regional traditions needs to be developed on objective
bases as well.
Basic stylistic characteristics were usually similar across the arts, as the
same stylistic norms ran through all the visual arts. Given their basic character
as a grammar of form, this could be expected. And yet sometimes significant
discrepancies occurred between different media even in the same art form.
Willett has shown that Zande (Zaire-Sudan) ceramics were more naturalistic
than Zande wood sculpture (Willett 1971:186). Skeuomorphism also must be
kept in mind. Thus, the patterns of brick decoration in Tozeur (Tunisia) and
nearby Algerian oases are clearly related to those of local rugs. Given the
geometry and angularity of the patterns, the wall decoration may derive from
textile patterns. Conversely, the architectural composition of the mihrab niche
and arch appears on textiles, in miniatures and even on scholars' slates. The
craving for innovations in arch design, so typical for the Maghrib, was related
to the decorative prestige of ever more sophisticated arabesques. The link is so
close that the arabesque decoration of the great Maghrebi minarets was
expressed as a series of intersecting arches with blind arches of the most varied
and fancy designs surrounding windows and even doors.
The use of architectural compositions to frame miniatures was quite
common in Ethiopia. South of the Sahara such mutual influences among the
visual arts have been far too little studied. Yet they are important, if only

Fig. l.lFagade. Tamelhat, eastern oases, Algeria. Bnck decoration, similar to those of the houses at Tozeur, Tunisia.
Decor directly (Berber carpets) or indirectly derived from textiles. After D. Hill and L. Golvin 1976:251

124

CULTURE AND ART

because they show how diverse was the common inspiration for further
innovation in the arts.
With regard to themes and motifs, the repertoire in a given culture was
normally quite consistent whatever the art form. But the distributions of

Hasan mosque, Rabat, Morocco. Height 30m. 1196, Almohadpenod. Decor differs
on each face of the lower. Compare with Plate 8.1, possibly by same architect

Plate 7. 1 Minaret. Stone masonry.

125

ART HISTORY

IN AFRICA

themes and motifs over the different art forms could be significant. Thus,
figurative painting was outlawed in North African mosques but not on
ceramics or metalware, nor at certain times as frescoes in bath houses, as
heraldic emblems and even as painting on glass for use at home (Atil 1981;
Talbot Rice 1965; Ayoub and Galley 1977). The distribution of decorative art
withm the Kuba corpus of art is intriguing. Two different ensembles of
patterns existed, one of which was reserved for the tattooing of women, the
decoration of a certain type of female skirt, and as decoration on the drinking
cups made out of buffalo horns, or their imitations in wood. The other set was
used in all the other contexts from wall decorations to mats, to ornamentation
of wooden objects, even jewellery and most textiles. Cases where the two
ensembles were mixed are extremely rare. Why this is so was no longer evident
by the 1950s to the Kuba themselves. Examination of the distribution of the
repertoire over the arts does in fact usually show restrictions of given themes or
even motifs to one art form or another. Such situations are not only of cultural
significance, but may also hold traces of historical development.

Fig. 7.3

Kuba

decorative styles. Embroidered textile

left,

geometrical repetitive patterns as on the right (this one

named) on

the

left.

vesstls. Left: after

VISUAL

The

latter type oj

Museum

decor

is

limited to

Tervuren 27401;

pile cloth right. All

Kuba

decor belongs

to

two types only:

the

and a juxtaposition of individual elements (all

women's scarifications, some women's skirts and buffalo horn

is

called 'rock')

right: after

African Arts, 1978, 12 (1): 27

AND PERFORMING ARTS

The relationship between visual and performing arts was often quite close
(Thompson 1974; Drewal 1980:18-20). Masks were usually designed for
display while dancing, as part of the decor for the performance. Public
buildings were designed for communal action, either praying as in church or
mosque, or for other purposes. The elaborately decorated niche walls of a

Swahili house, for example, were the backdrop to wedding rituals and
designed as such (de Vere Allen 1974b: 16-17). Plazas were made for meetings
and performance of singers, dancers, sometimes theatrical groups. The Ardra
board designed for use in Ifa divination implies the performance of the ifa
diviner.

Such relationships can

126

in fact be so close as to

become confusing. The

CULTURE AND ART

important difference between a mask and the costume with which it came, the
the mask was donned, and the dancing itself, is
the relationship to time. Performing arts need time. Their paradigmatic and
syntagmatic characteristics are separated. But visual arts were seen 'at once',
all their paradigmatic and syntagmatic relationships together. Yet ephemeral
visual art such as the body paint put on for one performance or even the mask

poems or songs chanted when

used during one initiation only, just as a complex hairstyle may last a season,
come close to being merely props for performing art. They differed from
statues in durable media. But even here clay statuettes or mud sculpture were
sometimes also designed to last for only one initiation (Cole 1969a, b, c). The
student of art must consider the temporal characteristics of the objects studied.
Were the objects designed to last? How long? Yet, even with the most
ephemeral visual creation, one must remember that they could be complex
expressions carefully crafted for an impact 'at once'. The affective impact of
the visual arts remains quite distinct from that of the performing arts.
Stylistic principles, or motifs common to both visual and performing arts,
were often hard to find. Themes were a common link often referring to myth or
other oral art such as the Chibinda Ilunga example or St George's story. In a
case, M. C. Dupre (1968, 1979; personal communication) showed that
the style of a two-dimensional round Tsayi mask characterized by almost total
symmetry around a horizontal equator as well as around the usual vertical axis,

famous

was made

for. In performance it was to be

should be the same upside down,
right side up, left and right. Moreover, she was able to link both dance and
mask to the upheavals of the 1870s, when the mask was created, and fortunes
seemed to revolve unpredictably. But this can still be explained as a feature of

was

directly inspired

worn by

by the dance

it

a cart-wheeling dancer, so that

it

use yet a use linked to performance itself.

To go further can be dangerous. Can one reasonably claim, for instance,
that the fioritures in Maghrebi vocal music corresponded to those in the
tracery of the lambrequin arches? Perhaps, but how does one prove it? As the
basic codes of musical and visual expression differ, what does such a statement
in fact mean? Ornamentation by deviation from the main development of basic

proportions? If that was meant, it could be tested both for musical and visual
styles. Were members of the community conscious of such analogies? If the
parallel works out, it then relates to secondary features of expression that could
be translated from one code of expression to another. Such parallels may reflect
similarity of use, or,

more

usually, reflect a

towards the perception of reality

itself.

But they are very

reality

by

common

approach

difficult to prove. It

simply to note correlations between the use of

different styles of dancing or music associated
with their appearance. One must, as Dupre did, show styhstic features of the
music or dance to correspond to similar features on the mask.
The complementarity of the different arts with regard to themes and
motifs in a given culture should be pursued in each case. The repertoires could
overlap or complement. Thus, while the graven image of man in Islam was
frowned upon and forbidden outright in places of worship, yet in poetry,
especially in love poetry, it was the main theme. The conception of man as a
grain of sand in God's universe was one extreme here, the other being man's
hubris in search of fame, power or love.
will not do, for instance,

different styles of

masks and

127

ART HISTORY

IN AFRICA

Plate 7.2 Maskers cartwheeling. Mask, wood, raffia, red, black, white colours. Dances now for fun. Likana, Congo.
Photo 1969. The mask, a disc, is symmetncal around vertical and horizontal axes. The designs around the face have
symbolic meaning. The mask was invented for the cartwheeling dance c. 1870

One can raise questions related to the dominance of any art form in a given
culture as well. It would be tempting to claim that the extraordinary
development of epics in the forest areas of Central Africa 'compensated' for a
relative paucity of sculpture, but if this may appear to be true for portions of
it was certainly not so for the western portions or the southeast of that
where both epics and sculpture flourished. It is difficult to state flatly that
Kuba sculpture, or architecture, or music or dance was dominant. All showed
similar sophistication and they were all linked to each other. Here epics were
clearly not developed but short lyrical poems had been brought to a high
degree of perfection. Can this be tied-in to a perception of the world also visible
in other arts? It is evident that nearby, in the cultures east of Lake Mayi
Ndombe, the main form of artistic expression was the ballet, from c. 1900
onwards. One can show there that villages invested great efforts in labour and
training to excel in that field. Such questions of interrelationship very quickly
force the inquirer to broaden the perspective to the whole field of culture.

the area,
area,

128

CULTURE AND ART

REPERTOIRE AND CULTURE

art expresses culture, so style, theme, motif are intimately linked to the
perceptions and the representations which are a common cultural good. In
general, the themes expressed in art are those dominant in a culture. Art
portrays only a tiny minority of shapes and objects perceived. South of the
Sahara for instance the repertoire in sculpture consists first of people but
includes also some animal representation and almost no plant life, while some

As

plants are the foundation of a whole stream of arabesques north of the Sahara.
The absence of animals could be linked to notions of inert life as opposed to
moving life, were it not that only very few species of animals were usually

few favourites such as elephants, buffaloes, leopards, spiders,

come to mind. In studying an animal repertoire, the first
error is to assume that the classification of animals is the one we are familiar
with. Zoological classification had nothing to do with representation. The
animals portrayed were all important symbols in the culture (Ben Amos 1976).
They were 'good to think and to feel', apt metaphors to express profound
thoughts. Not all crucial animal symbols were necessarily represented either.
Thus, owls and similar birds were considered in wide portions of Africa to be
omens or symbols of witchcraft. They rarely occur in art. An inventory of
representations should be established first for a given culture, then the
significance of its content can be established with regard to the whole culture
drawing not only on performing arts or oral expressions but on ritual, daily life
and so on. The whole texture of culture can be brought into play, to reach a full
understanding of the collective representations that were expressed in art.
Themes could not only be selected but rejected. We cited Islam's ban on
the graven image as an expression of idolatry. It was almost totally effective in
places of worship and most strictly enforced with regard to sculpture, but
elsewhere the ban was not total. Bravmann has shown that masks which once
had religious connotations continued in use among Muslims of northern
Ghana or in Nupe (northern Nigeria). Indeed, masks for fun were still to be
found in nineteenth-century Tlemcen (Algeria). South of the Sahara we cited
the common avoidance of portraying owls. Witchcraft and misfortune were
generally not to be represented at all, so no witches or diseased persons, no
slaves to be traded appear in most arts south of the Sahara. There were
exceptions: in the arts of Ife and along the Cross River (Ibibio) as well as along
the Middle Niger before A.D. 1500, diseases were shown on masks or statues,
probably to illustrate the powers of the entities that caused them.
Unfortunates, slaves and prisoners appear in the an of Ife (Eyo and Willett
portrayed.

lizards, crocodiles,

1980; Willett 1967).

Evil

and the

effects of evil are rarely represented.

The Lega who had an

aesthetic of the ugly associated with evil are an exception (Biebuyck 1973:

ill.

Witches were apparently never portrayed as persons. Perhaps

representations of these aspects of life were suppressed because images become
real and power might emanate from them. That is the reason given in
Ethiopian painting for the convention always to represent Judas with averted
eyes; his gaze, the evil eye, could hurt the beholder. In many cases, however,
the identification between beauty and the moral good coupled with idealism in
66-7).

art explains the situation sufficiently.

129

ART HISTORY IN AFRICA

The

repertoire gives a profile of the

dominant positive ideas, ideals,

whether they be

beliefs, values, gods, spirits, ancestors, mystical forces,

religious or not.

Some

representations also reflected poetic but anecdotal

Sometimes depictions of sacred history appear, as on the walls of Faras

or as on Dogon (Mali) sculpture, sometimes allusions as in the representation
of the thunderbolt of the Yoruba (Nigeria) god Shango, that appears in the
shape of a double axe. The frequency of the appearance of a theme and the type
of object also tells a tale. Richly decorated utensils stem from courts, or denote
scenes.

special significance in other cultures.

The

only artistically treated objects

of northern Congo are their honey adze, the

shafts of spears, the back pack and bark boxes. Honey is the most desired food,
and once a year the search for honey broke up camps and led to a different
pattern of life. The other objects are equally important (Demesse

among

the

Aka pygmy hunters

1980:113-15). Decorated granary poles, doors or locks in the Sahel

emphasized the importance of storage even if, among the Dogon, they also
expressed the meaning of the granary as the ark of humanity. Thus the
inventory of the visual arts in any culture tells us what its main concerns are,
barring those that it was forbidden to depict. Even tourist art, when not wholly
imposed on the producer, expressed cultural and social concerns.

Nineteenth-century Loango tusks depicting the procession of

a caravan, the

settlement of a court case, the carriers offish, the local lords in their finery, the
foreign traders with their barrels, chests and padlocks, are microcosms of what
the carver perceived Loango to be. Many a scene could not really be

understood by the European buyer, yet they were carved as part of the
statement about Loango. Now they are reminders of the customs and the
times.

AESTHETICS
Western aesthetic criticism of African art is largely irrelevant because it is an
expression of Western culture about what to them are objets trouves. Yet
Western analysis of form can correspond to the analysis made by the artists
created these works. In that case the art historian stil' needs to be
informed about the aesthetics of the culture that created the objects. The
agreement between Western aesthetes and their local colleagues may be a
coincidence of taste or the mutual recognition of explicit formal values.
Pronouncements about 'good art' and 'bad art' reflect Western preferences to
the extent that some anthropologists have claimed that 'African art' does not
exist at all. It is but the study of Western sensitivity towards African objects
(Maquet 1979). To the extent that aesthetic analyses are simply this, they are
indeed spurious. So the very first requirement for any study of aesthetics is to
know what the aesthetic criteria of appreciation were in the cultures from

who

which the works emanate.

Until quite recently, however, such studies were almost non-existent, and
a critical reading of the ethnographic information. To
claim that objects were made for 'art's sake' in places where tourist trade for

some do not survive

colonial officers and traders was common is spurious. To claim that criteria can
be deduced merely from a few interviews with random viewers asked to rank
objects by 'beauty' is unconvincing. First, such beauty contests did not often

130

CULTURE AND ART

to tourists. Loango coast, Congo/Angola. Wallers Art Gallery, Baltimore. Height

10cm. 1850s. Such tusks were carved from the 1830s or 1840s to c. 1900. The practice spread from Loango north as far
as Duala and south to beyond Luanda. Scenes represented on the photo: caravan, European traders, sale, African
notables, men and women, execution, ape. Other scenes of the tusk show war, capture of slaves, fishermen, a carpenter
working at a table and again the ape motif at the bottom. A slave caravan is also depicted. The composition is European
inspired, but the execution and the interpretation of each theme is local, and carries perhaps implied symbolism. We see

Plate 7.3 Carved tusk. Use: sale

1

the

mid 19th

c.

through Vili eyes

131

ART HISTORY

IN AFRICA

why one

item was thought to be better than another, nor did they

pubUc at large and the creators of art. Only statements
by the latter would really be detailed enough to allow us to understand the
principles of aesthetic criticism involved, and fmally some contests did not

explain

distinguish between the

even make a distinction between works foreign to the critic and works crafted
by him. Any 'beauty contest' should take great care over the sample of persons
interviewed and include artists, while the relationship of object to critic

should also be clearly specified (Gerbrands

(maker, owner, user
.)
1957:75-8, 89-90).
There is no doubt that art objects were first appreciated for their
emotional appeal, at least by the general public. That appeal derived from their
use, meaning and place in the social matrix. Thus, a type of Cameroon
(Bangwa) mask was seen as terrifyingly beautiful because the members of the
'night' society wore it. This was the chiefdom's secret council holding power of
life and death over all. Its members secretly executed the condemned. The
mask type represented terror (Brain 1980:147-51; Harter 1972), and its
.

aesthetic success
feeling.

depended therefore on

The type complemented

its

expressionistic

that of the 'gong' society

power to evoke that

which expressed the

protective aspects of the chieftaincy.

When
first that

comparison

starts

with icons of the same type, it becomes evident

By working with carvers from the

aesthetic criteria indeed existed.

westernmost Ivory Coast (mainly Dan), Vandenhoute was able to specify the
following criteria: symmetry according to vertical axis was consciously sought
and strictly maintained, but there als3 had to be balance, rhythm and harmony
between the various volumes, surfaces and lines in the mask. Such concepts
seem vague, but in fact are not and can be measured. Apart from such formal
criteria, finish (polishing and staining) was taken into account as well as criteria
of actual use such as the comfort of the wearer of the mask (Gerbrands
1957:91). R. Thompson's (1973) study of Yoruba artistic criticism found
eighteen criteria of sculptural excellence, each being named abstractions, tor
the Yoruba use a technical vocabulary of aesthetic criticism. Most of these were
formal and once the criteria were
evaluate sculptures in a

way

known it was easy for the Western art critic to

similar to the

Yoruba. However, they also

all

related to larger cultural values, to a system of ethics. Hallen (1979:303-7)

shows

that

Thompson's claims

are exaggerated, but, equally important,

even though some are inarticulate. It is to be

deplored that we still have only a handful of valid studies and none dealing with
the arts of northern Africa. To a certain extent aesthetics can be deduced here
from writing, but field studies would no doubt enrich the understanding of

basically the criteria exist,

if in future such studies become available in larger

be lacking with regard to all the art forms that have now
died out. One can only hope that it will become more and more apparent that
formal criteria deducible from the works themselves will in fact turn out to be

such aesthetics. Even

numbers, they will

still

the intended criteria by

which they were made.

If this

were

a general rule

sense would claim that it is so, but common sense is often common
prejudice - then formal study could, at least to some extent, palliate our total
lack of information about older art.
saw that the link
In many cultures 'beautiful' was rendered as 'good'.

common

We

was so strong
132

in

some

cultures that 'bad things' could not be expressed in art.

CULTURE AND ART

But the notion of 'goodness' or 'beauty' or

'quality' varied

from culture to

denotes 'approval'. The cultural conditions in North Africa drove

aesthetic expression to the line, to arabesques. Artistic sensitivities ignored the
sensual attraction imparted by volume but revelled in the mathematical
elegance of intersecting lines and in the play of light and shadow while prisms
of colour, if any, merely accented the subtlety of the play of lines. In contrast
culture:

it

art south of the Sahara expressed volume and the sensual, the solid rather
than the evanescence of light. Some here were attracted by the clash of angular
volumes as in eastern Kasai (Songye) (Plate 5.6) or in eastern Liberia
(Guere/Wee), some prized smooth transition above all, deriving compositions

most

from the smooth sphere

as in

Shaba (Luba) (Plate

5.4) or

from

elliptical

volumes, swelling rhythmically as in northern Gabon (Fang) (Plate 5.3). Some

relished rich decoration, glossy lacquer-like finish, while others liked the stark
simplicity of the unadulterated medium.
Can one then posit links between the specific aesthetic expressions and
other cultural values? Would smoothness correspond to gentle behaviour and
clashing volume with martial sentiment? Are the lovers of arabesques and
light, rational analysts or deep mystics? Is hieratic art the expression of
religious awe or of an appreciation for authority, dignity, tradition? Such
broad links are always false guides. One would not expect everyone in a given
community to be gentle, martial, authoritarian or docile, pompous or affected.
Styles are no better indicators of the overall tenor of a culture than languages
are. Sweet Italian and harsh Russian, like sweet Swahili and guttural Arabic,
are but stereotypical prejudice. Styles are partly autonomous from the rest of
culture insofar as they are arbitrary systems, learned like language in
childhood. If links between beauty and the tenor of a culture are to be explored
fruitfully, they must expressly be found probably through the analysis of
aesthetic criteria and their link to ethical criteria in a given community.

DYNAMICS OF ART AND CULTURE

and convention were so important in art, how could change occur? In
way as culture as a whole changed and in the way language changes.
People in a village speak a dialect, but each person has his or her favourite
expressions, words and turns of syntax -each person speaks an idiolect. In the
same way, concepts, visual concepts, the perception of style, were generally
held in common, but each person has variants of his own. Artists, more
conscious of their visual heritage than others, were bound to explore it more,
just as poets use the language differently from ordinary people. Such
idiosyncrasies allow new expressions of art when innovation was required by
social or cultural change outside of art.
In most cultures adherence to norms was appreciated as the familiar, but
at the same time it was not enough. Within the general framework of tradition,
innovation was called for. New variants on the same theme in the same manner
were highly appreciated in all the arts. As long as the new creations did not
deviate so much from the current canons that the strain of understanding the
icon was more unpleasant than the appeal of novelty, innovation was
encouraged. Artists thus educated their public by finding new variations on
old themes, old volumes, old lines, old colours, old motifs. The only surviving
If style

the same

133

ART HISTORY

IN AFRICA

record of these processes are the art works themselves and the change recorded
them for the last century and more.
The distinction between oikoumenical and regional tradition relates also to
the whole culture and the way that changed. Oikoumenical visual concepts were
widespread. Everyone in the oikoumene did not share everything but most

in

common cosmologies, common rituals, some

common institutions and common values. Literate persons shared much more
persons everywhere shared

than that with other literate persons elsewhere, but not necessarily with their
diverse and
illiterate neighbours (Goody 1968:1-68). Culture was more
in
architecture
have
we
stratified in the oikoumene, and so was art. Thus
find
yet
and
accent,
local
a
with
tastes
Muslim
general
Morocco that reflected
of
rural carpets there that were not replicated anywhere else. The ceramic art

Kabylia in the nineteenth century was really unique, whilst the oratories there

were commonplace.
Regional traditions usually lacked such great differences between the
of different strata. Because literacy was absent, they lacked
overarching institutions and world views. As a result, their arts were less open
but from time to
to mutual influence and did not coalesce into a larger stream,
art spread with
and
spread
ideas
and
values
associated
the
and
time institutions
culture

them. For example. Cross River styles were carried by the Ekpe association
along with trade almost to the full length and width of the valley (Eyo 1978).

Plate 7

A Mam enclosure from the air, Zimbabwe. Sione masonry without mortar. Diameter 135m.

early 15th

c.

Centre of the

from perhaps 1000

134

to c.

Zimbabwe

1830s

state.

An example

of stone architecture

Outer wall 14th or

which developed

in southeastern Africa,

CULTURE AND ART

Loango styles spread to southern Gabon along, no doubt, with the
expansion of the kingdom. Royal architecture in stone at Zimbabwe spread
over the whole country of Zimbabwe and remained quite different from most
local architecture between c. 1200 and c. 1450 (Garlake 1973). The river of
style out of which Zimbabwe grew may well have begun further south on the
Limpopo. But once it became a dominant court style associated with court
culture, it remained a hallmark of kingship in Zimbabwe to the eighteenth
Naturalistic

century. Meanwhile, further south, in the high veld, architecture in stone

developed in different directions and, under the influence of cultures that
diverged more and more, it too diverged so that a so-called Tswana plan of the
eighteenth century was very different from plans in the eastern Transvaal or
State (Inskeep 1979:135-41).
not difficult to give other examples both of the expansion or
diversification, or contraction of art styles linked to similar general movements
in the cultures these arts expressed. Art is an important and direct signal of
culture change and hence of general cultural history. This is evident even in
such seemingly minor changes as the dominant colours of frescoes at Faras.

Orange Free
It

They

is

reflect

renewed influence from

pulse of cultural

life at

the cathedral.

meant more Palestinian influence

orientation.

More Byzantine

different quarters

and changes

in the

More
in

Palestinian influence in art also

reading matter, and theological

influence implied

more use of Greek

again,

more

attention paid to Byzantine church fathers and their theology (Michalowski

1967; Dinkier 1970). Any historian of culture must therefore be alert for
telltale signs in art, because they may indicate the possibility of less visible
change.
Sources for the history of culture are available through writing. Those
that spurned this medium of expression left little record of cultural change. It
is ironic that we know more about cultural changes in the New Kingdom of the
Pharaos than we do about eighteenth-century Kuba cultural change, three
thousand years and more later. Perhaps the greatest weakness of precolonial
African history south of the Sahara is the lack of data and study of cultural
matters. I still do not know whether we will ever be able to trace even the main
outlines of a history of cultures in the sense in which the term was used in this
chapter. If this can be done, works of visual art will be the most direct evidence
we have, and it should be given the most careful attention. In the end, a study

of art history might therefore well lead us to a better understanding of cultural

history in general.

135

CHAPTER EIGHT

THE CREATIVE PROCESS

THE ARTIST AS CREATOR
Having examined questions of authenticity, identification, context,
technique, style, iconography and meaning, we now ask how an icon comes
into being, what were the sources from which it sprang and how original is the
work itself. The creative process starts both with perception and cognition.
Perception of an artefact, a person, a natural object or scene, is probably
common to all people everywhere, although Gombrich doubts even this.
Cognition through mental predispositions of space, volume, number and time
is common to all people in a culture. So are common expectations and
interpretations of perceived phenomena. In addition, artists have further
developed interpretations linked to the task of translating an object into a new
work of art. Artists know from practice. They have learned how to translate
perceptions into works of art or can see the problems involved in so doing,
especially when they see icons in media and techniques which are their own.
Apart from such technical predispositions, however, most cognitive factors
were shared by artists and non-artists alike in the whole community. There
existed, for instance, an almost universal measure of agreement as to what was
worthy of observation and transmutation into an icon.
Perception and cognition were conditions of observation. Motivation
the
of
image
mental
A
action.
for
ingredient
essential
-was
the
the will to create
icon related to one or more material images to be made had to be present in the
mind of the artist and his patrons. The mental image of the artist then
purposively directed his skilled hands at every step of the crafting (Drewal
1980:9-10). Once the icon emerged, it escaped from the artist's control. As a
reality in its

own

right

it

started a

life

of

its

own.

be proved in the absence of express statements by

artists or patrons. Even so, motivations are always mixed: practical reasons
accompany and vie with more aesthetic urges. But the product clearly bears
witness to the mental vision of its maker, especially when the artist made
replications, copied or attempted to perfect or adapt existing works of art.
Even when new icons were created, the product remained within the range of
Motivation

itself cannot

earlier techniques, styles,

The

themes and motifs known

around three main

creative process turned

to the artist.

pivots: the patron, the

artist, and the products. However important the influence of the mental
images of patron and public were, they only set limitations or requirements on

was the person who turned a mental

was crucial. In most cultures this was
more or less recognized. Creativity was attributed in most cultures south of the

the creation of acceptable

art.

image into matter. That

why

136

is

The

artist

his vision

THE CREATIVE PROCESS

C
Prime work

Reolica

Artist's

Cognition

work

Icon

Mental image

Perception

Other

Replica

etc
Fig. 8.1 The creative process

Sahara to supernatural forces which made visible out of invisible matter.

Prayers or sacrifice to media, such as metal, wood or stone, to the artist's tools,
by the
to the processes, have been reported. Ritual conditions to be observed
life of
invisible
an
hold
to
were
artist, especially when making objects which
137

ART HISTORY

IN AFRICA

own as supports or even images of spirits, were fairly common (Drewal

1980:10-11). Skill alone was not a sufficient explanation for creativity. There
was nothing surprising about this since skill in anything was linked to the
supernatural, often unfavourably: people suspected that any outstanding skill
was acquired through witchcraft or sorcery. In many societies inspiration itself
was claimed to stem from dreams by patrons about commissions specifying the

their

icon desired. Such an exotic picture of creativity can and has been overdone in
ethnographies. Just as not all African art is inspired by religion, neither was all
creativity inspired by prophetic vision nor was it totally attributed to it. But
creativity was understood in a framework of the dominant cosmology as a
matter of course, and this is equally true in Christian or Islamic countries.
Muslims not only heeded taboos about representations, but like Christians
attributed creativity to a condition of special grace (baraka) and not just to

Religious commissions were more common in Ethiopia than elsewhere in

Most of the painting and building there were the work of monks, while
only a small portion of carving south of the Sahara was done by religious
specialists. What matters more than such rather invidious comparisons is the
skill.

Africa.

realization

that

motivations

for

making works of art, and attributing

and success in their material crafting,

responsibility for both mental images

were embedded

in

cosmology, but did not exclude appreciation of the

individual input of the artists.

was recognized everywhere but was not easily visible to the

works were seldom signed, even where writing existed. For
every few Coptic or Muslim potters who signed the scenes they painted on
Fatimid bowls, there are legions who did not. Patrons are mentioned on
inscriptions on Muslim monuments, but seldom architects. Western Liberian
(Gola) patrons were also remembered, not the artists. In our age we have
deduced that this means that both the public and probably the artists saw such
realizations either as communal work or saw no special merit in crafting them.
But this is not altogether true. For some names of painters in Ethiopia or
architects in northern Africa were remembered in writing.
South of the Sahara the situation may have been similar. Signatures do not
exist even where marks do, and we must remember that marks were signs of
ownership, not of manufacture. In some societies, such as among the Kuba or
the Yoruba, innovation was prized. The invention of a new decorative motif
among the Kuba was rewarded by calling it by the name of the maker. Thus,
the motif Mabiinc, later quite common on mats, bears the name of the woman
who invented it. In most communities, artists could tell from the manner of
execution who had shaped this or that piece, and among some, as in western
Nigeria (Yoruba), even the public knew who had crafted what were thought to
be significant works. The particular mix of attribution to inspirations by
supernatural forces and inspiration by the artist varied from society to society
and certainly varied over time, but individuality was never denied, certainly
not by artists. But the conditions of collection have been such that most of this
information was never recorded.
The artist was not just considered to be a hand or a tool. In most societies
Individuality

historian, because

he enjoyed consideration,

West

Africa,

for

not status.

instance,

Consideration could

138

if

creative

be expressed in

Among

the social castes of artists of

was ambiguously received.

twisted ways. Thus, in western

skill

THE CREATIVE PROCESS

Cameroon, good artists were eagerly traded from chief to chief as slaves, and in
the Bangwa chiefdom there, they worked around the chiefs palace and ceded
their products to him. He sold or distributed them and claimed the credit for
the inspiration! The chief was the official artist (Brain 1980:135). Artists
appear very rarely as heroes in oral literature; only among the Kuba do I know
of a tradition glorifying a member of the royal house as the creator of later lost
art works of great size in iron. But appreciation for artists has been preserved in
many societies by the characteristics attributed to demiurges, who 'invented'
arts

and

crafts.

was recognized in most, perhaps all societies. Did this

recognition involve the attribution of special roles and peculiar behaviour
patterns to artists? Was there an 'artistic temperament'? Bohemian ways of life
and unconventional behaviour have been reported for musicians in eastern
Creativity in artists

Kasai, but not for visual artists there (Merriam 1973). While in our societies
this feature is linked to the notion of artistic freedom, a right derived from the
creed of art for art's sake, it could exist in societies where creativity was linked
or
to supernatural inspiration and artists could be likened to mediums, healers
mystics. There are African societies in which those people behaved differently
either during a given period of their life, when the vocation called them, or in
seances. Artists were expected to play special roles in some societies, roles
quite similar to those of inspired persons like mediums. In such cases it was
believed that a special bond existed between the artist and a supernatural being
by which he was possessed. d'Azevedo (1973a) says of Liberian (Gola)
woodcarvers that a tutelary spirit was held to be the source of their creative
ideas. Already in childhood deviant behaviour could be detected. Signs of
strong spiritual connections, unusual food and preferences, early attempts to
master skills, were all signs of calling as a woodcarver. Their behaviour in later
strongly
life was seen as strange, a mixture of the irresponsible and the
committed to goals that were not important to others. But there also are
African societies in which artists learned their crafts just as farmers learned to

farm, hunters to hunt, or weavers to weave.

Few studies have focused on actual temperaments as opposed to roles.
They could be expected to vary as much as the range of temperaments in the

whole of their community. Thus, among the Kuba, we found one well-known
carver to be an advisor and tax collector for the king, a careful person who
budgeted his time and calculated his output, a consultant to the local art
school. Another, living early in this century, was driven by ambition, highly
talented, frightening to others and competitive. He was killed around 1904,
framed by jealous competitors or villagers convinced of his evil powers. He did
not survive a witchcraft poison ordeal. A third carver I knew in the 1950s was
the perfect trader.

He carved

a statue,

went

to

town

to sell

it,

calculating quite

how much work, and how long the money

nicely how much it would
Another man of cheerful disposition
shoddy.
not
was
work
his
last.
Yet
would
was engaged, when I met him, in displaying a new variation on an old theme to
his fellow villagers. Carving was common among the Kuba but an artistic
bring for

temperament was unknown, and

from the Gola.

and actually played, vary widely from

in this they obviously differed

Clearly, variations of role, expected

culture to culture.
The degree of freedom enjoyed by artists in creating

is

linked to the role

139

ART HISTORY

IN AFRICA

they played, the behaviour expected of them. Thus, the Christian monk who
painted murals from a pattern of a copy book on set themes and in specific
places of a church had very little freedom. But the question of the freedom of
is partly spurious. Few artists ever perceived limitations on their
freedom of expression, as they themselves were embedded in the
collective representations that produced icons and were usually inspired by
signs and motivations common to all. However the question is partly genuine.
The monk painting frescoes had less freedom than a carver replicating existing
icons, and yet, as the record shows, innovation occurred. From time to time,
startling departures from conventional work did occur, always of course within
the world of forms known to the artists, not the larger world of possible forms.
When discussing artistic freedom it must be remembered that the
contemporary artist has access through illustrations and museums to forms
shaped the world over and at all periods, quite a different experience from the

the artist
creative

usual situation in the cultures of Africa.

The creative urge as a motivation is a vague phrase. It includes

motivations such as boredom, the wish to fmd a better solution to a problem of
form, the desire to record permanently, the desire to make a statement.
Evidence from series of similar icons often shows minor variations that may
well represent the urge to escape from the boredom of executing the same icon
once again. In a few cases variation may be motivated by the desire to resolve a
formal problem better. This was a major source of innovation in European
graphic art and architecture. The treatment of the wall behind and over the
mihrab

in Moroccan mosques is an example of an architectural conundrum.

Different solutions were tried in different mosques until a very satisfying
solution was found at Tin Mai (Margais 1954:238^2). That became standard

practice thereafter. But given the lack of expressed motivation by artists we

cannot often be certain that this incentive was at work. Clear cases of such
solutions found after various trials are practically non-existent because dating
so poor

and

often tend to put the best integrated work first in a

whereas less perfect works may have preceded them as
trial pieces. The desire to put on record is clear when portraits were crafted,
especially portraits of those who were near and dear to the artist. Examples - all
of the twentieth century, however - are known from various parts of West and
Central Africa. As to the desire to make a statement, such as Picasso's in
Guernica, the evidence is ambiguous at best, in the absence of literary
indications. As we know from oral art, there must have been cases when this
was the motivating force, but we cannot detect them.
When artists departed from current work they claimed credit for the
virtuosity of execution, and for the thought involved in turning new mental
images into reahty. Existing ethnographic reports often stress the claims that
thinking was the hardest part of the job and execution the easiest (Gerbrands
1957:124). Equally, often both works and reports stress virtuosity for skill's
is

stylistic series

series, as its prototype,

sake.

The

latter incentive led

to the creation of bravura

pieces, usually

non-commissioned, which were one of the major means of innovation in the

arts, and, at least among the Kuba, an accepted form of advertisement for the
artist (Plate 8.5).

140

THE CREATIVE PROCESS

ORIGINALITY
question to be asked of every object, as of every document, is its degree of
may be perfectly genuine - that is, not pretend to be
anything else than what they are, but they may still not be original. Authentic
pieces can be copies just as they can be prime works. The notion of copy in art
depends first on the techniques used. Several items from the same mould,
several prints from the same block, are true copies. If the same pattern is used

originality. Objects

to trace the outlines of a figure for a fresco, as

and Ethiopian
presuppose a

was probably common in Nubian

we have copies, but not total copies. Such mechanical copies

different original: a mould, a block, a pattern book. The

art,

to reproduce copies should be well known. Unfortunately,

any of the moulds and no blocks or pattern books seem to have

technical

means

very few

if

V4

V4

V4

V4

e*

Fig. 8.2Mihrabii;a//. Tin Mai, Morocco. Stone masonry.

floor-level passes into a bipartite one on the top of the wall.

153 (Almohad period). The horizontal tripartite division at

The

mam

transition area counts seven units. Vertically

two with each half subdivided differently, the upper part m equal thirds, the lower half one-third to two. A'
corresponds to A; B'toR. All dimensions are related to that of the diameter of the arch. The composition developed here
became standard in many mosques thereafter
division in

141

ART HISTORY

IN AFRICA

survived in Africa. Pots can be copies in yet another sense. Where the pottery
exist the bottom of an old pot often served to mould the bottom
of the new one, so that the size of the new pot and most of its shape (given the
colombine method) were dependent on the model.

wheel did not

Often the term 'copy'

replication.

is

used in a different sense altogether.

It

means

pattern, a design, a sculpture, a building, were crafted in

conscious imitation of an existing model. A marabout's tomb in northern

Africa resembles similar tombs to the point that one talks of a copy. One statue
looks very much like another one, as anyone who has seen collections of twin
figures from southwestern Nigeria (Yoruba ibeji) or stocks of tourist art with
Masai warriors wielding spears knows. But such replications are not total.
They vary slightly in size, in technique, in the grain of the medium, in a detail
(very common in painting and architecture) or they are composite copies,
using one figure for part of the composition and another for the other part. The
many St Georges and St Michaels with varying dragons and fallen angels in
eastern Christian art illustrate this well enough.
Nearly complete handmade imitations were the rule in preindustrial
material cultures, for example in textiles, utensils in wood, metal or pottery, in

house building. The confidence archaeologists have in determining cultures by

their pottery or their house plans underlines the point. Would art objects then
be those that differ from run-of-the-mill production? Or perhaps just the
prototype? Obviously such a criterion leads us into a quagmire. The art
historian merely places the object he studies from the point of view of its form
in relation to all others. Series of replications do not rule out that there is a work
of art here. Most African masks are taken to be works of art, even though
hundreds of replications exist for many of them.
Originality can also be interpreted in a third sense, referring to creative
filiation. A sequence of works established by consideration of their similarities
yields a time series. Some works are then labelled as archaic, some ds copies,
crafted by imitators who aimed at emulation or even improvement. The use of
such terms presupposes a whole hypothesis about artistic development. Given
the often undated character of works in Africa and the unproven assumptions
of a wish better to solve technical problems posed by one's predecessors, such
terms as 'archaic' or 'epigones' are best eschewed. Classical is used then for all
art before c. 1900. A terminology using 'replication' and 'prime work' is
adequate and less loaded (Neyt 1979:24-6; Flam 1971).

REPLICATIONS
Most works of art
works were strong models. They were imitated
through a mental screen, the image of the icon to be imitated. Familiar motor
habits and usual skills crafted the object.
Working from nature was as rare in Africa as it was in Europe before the
Renaissance (Gombrich 1960:7483, 148-52); not only unusual, but distinctly
odd. Hence counterfeits, exact reproductions of existing objects, were rare, if

Creative processes led either to replications or prime works.

are replications. Existing

the techniques did not provide for counterfeiting.

Exact reproductions
required continual inspection of the model and imitation detail by detail from
the model to the new product. Such exercises were totally foreign to the

142

THE CREATIVE PROCESS

who learned by imitating the actions of their
mentors until they gradually mastered the whole sequence of operations and
motor habits required to fashion the products of the workshop to which they
were apprenticed. Practising artists might keep some objects in their atelier,
either because they were not yet finished, or not sold, or were rejects or in order
to serve their memory. Some were mementoes to great occasions in the artist's
life, some works he liked so much he would not be separated from them. I
knew one Kuba carver who held on to perhaps a dozen objects but I cannot say
that even one of them served as a model. Given these conditions working from
memory never led to lifeless creation or total imitation. There was always room
training of African artists,

for innovation.

The ability
and those

to distinguish

that are original

between those features

is

work that are copied

The determination of

in a

crucial in art history.

personal styles, filiation of art works, all depend on the skill with
which the origin and derivative qualities of replications can be analyzed, and
replications differ between major art forms.

ateliers,

Almohad
Plate 8.1 Minaret. Stone masonry. Quttuhiya mosque, Marrakush, Morocco. Height 77m. 1146-1196,
Compare with Plate 7.1. Decor is here more functional, hut less logical than on the Hasan minaret

period.

143

ART HISTORY

IN AFRICA

In architecture, circumstances usually made it impossible to work with

same plans on the same scale. Replication meant to use the same principles
for plans and elevations, the same techniques, the same standardized
materials. A famous example of three minarets replicating each other, built
reputedly by the same architect in Seville, Rabat and Marrakush, illustrates
the point (cf. Plates 8.1 and 7.1). They all varied slightly in dimensions, but
not in the method of construction. Interior ramps, not staircases, were used,
and there was no central pier supporting the whole tower. Each level
constituted a room. The three towers finished in 1196, 1196 and 1198 all
differed from previous minarets in this feature. They all shared the fact that
each of their four fagades was decorated differently. They all stand as a single
group different from earlier minarets and were models for later towers
(Margais 1954:209-11). Within the group further differences were the
repartition of the decor at Rabat which was less functional than the decor at
Marrakush and the variations in the slope of the ramp at Seville allowing a
better alignment of outside windows. The delicate balance between the desire
to innovate and the desire to be faithful to the model the patrons wanted is
the

evident here. So is the architect's experimentation with technical solutions of

decor and interior ramp. One of the main differences between these minarets
and earlier ones was the technical innovation that allowed suppression of the
central pier.

Additive techniques in sculpture showed fewer limitations in obtaining a

The material was more homogenous. Even with
new moulds or with any of the other techniques used, the imitations could be
quite close because the building-up process allowed greater precision. Use
could also further imitation. The need to control weights in the manufacture of
jewellery and goldweights led to closer imitation of size than in other arts for
instance (Garrard 1980). Standard sizes of painted tiles had much the same
effect. However, because it was possible to imitate so well, individual hands
are much harder to detect in ceramics or in metal work, and variations between
one work and another by the same hand were often greater than variation
between replications by different hands.
Subtractive techniques led to lesser replication. The shape and features of
close imitation than carving.

the initial block, the techniques of carving prevented very close imitation.

The

use of adzes, chisels and knives preserved more traces of the hands of
individual carvers and thus enables identification of products from the same
atelier and by the same artist, but polishing and finishing with the knife could
erase so many of these individual traces that it becomes impossible to recognize
the hand. Hence, arguments of originality in subtractive sculpture need to be
realistically assessed in relation to the objects involved. As a rule of thumb, the

more

finished the work, the less convincing such arguments are, the rougher

more plausible.
With two-dimensional arts, especially

the sculpture, the

painting, other questions arise.

Tracing apart, differences are found in execution. The examination of

'signatures'

is

certainly fruitful here.

For

just as graphologists recognize

different hands, the art historian finds the individual characteristics of how the

paint was put on (see Fig. 5.4). Tracing did not occur for illumination nor for
the painting on wood, freehand imitation preserved all the characteristics of

the atelier and the master. But copying illuminations was often quite

144

literal,

THE CREATIVE PROCESS

just as

copying a manuscript

is.

It

leads us to think that while scriptoria

of the individual hand may be more

canvas
(but see Leroy 1968:65-6).
or
difficult than in painting
Similarly, the production of other art forms, such as textiles or work on
leather implies greater or lesser imitability. Decorative art was often easier to

(ateliers) are easy to identify, the detection

on wood

copy from memory than figurative work was, and hence

individual characteristics.

Any

it is

harder to examine

historical critique of originality

must

carefully

take such factors into account.

DRIFT
Because replications were never total copies, replications of replications tended
to drift away from the first original that served as a real or mental model.
Further, in time this drift could lead to stylistic change. Drift could be random
or not, and random drift was very common. Small mosques in the Sahara, for
instance, differ in the distance between arches, their actual heights, the
execution of the roofing, and the internal proportions of the ground plan.
That, however, did not lead to definite change over time, as the variations were
considered to be unimportant, occurred in various ways, and tended to cancel
each other out. Random drift in other arts is equally common. Drift became

and is of major concern to art history when it was directional. The

developing from the garlands of vines, fruits and leaves
arabesques
history of
in Hellenistic decoration through a gradual stylization leading to geometry is
one of drift. Arabic script of the Kufic style began as a monumental austere
script, but later sprouted leaves and became flowery Kufic. It became less
stylized in the same workshops where the rinceaux became more stylized at the
same time. The example shows that there are no laws involved in drift. It was
strongly believed at one time that decorative arts simplify and go from more
lifelike representation to simplified forms, presumably to speed up execution
and to bring out the essential characteristic of line in the representation. But, as
flowery Kufic shows, a reverse trend was just as plausible. Later on, further
development of Kufic turned it into such an angular style of decoration that it
looks like clusters of little blocks. Over a period of almost a millennium in

significant

duration, the full evolution shows a movement away from simplified line to
more Ufelike representation and then back to a very abstract play with surface.
Directional drift occurs when artists develop a personal style, when they
seek to improve on existing icons, when they want to produce similar icons

and when taste changes. Individual styles have been documented, if only
by documenting the hand of a painter or carver, or when works from different
stages in a career are compared (Thompson 1969: 120-82). There is always an
faster

interaction

between patrons and producers

in this process.

The attempt

evident in architecture and we cited the case of Tin Mai

as an instance. Change in order to speed up production is proved by shoddy
work and best known by modern replication of classical icons to be sold to
tourists, but this factor may have played a greater role than is usually
formally to improve

is

acknowledged. Certainly this is true in the development of hieroglyphic script,

and we suspect it in many other cases of simplification of decorative motifs. By
and large, change in taste however accounted for most drift. A replication was
deemed to be superior to the original in some detail that had changed more or
145

ART HISTORY IN AFRICA

by accident. The next patron expected the next icon to display this
alteration more prominently and gradually the whole composition or style
changed, sometimes quite considerably. Thus, the Ethiopian keyhole window
was the endpoint of a drift starting with a window flanked by colonettes on a
base supporting an arch on a corbel. A series of Tyi wara masks from Mali
(Bambara) shows all the transitions between a flowing mane over a majestically
arched neck to a flamelike shape in which no mane could ever be recognized.
less

Fig. 8.3 Tyi wara masks worn on top of the head. Wood. BamharalBamana, Mali. Recent. The objects all represent
male antelopes. The styles vary by re^on and over time. C consists only of the horns and mane of form A. A Bamako
region: height

40cm and

the best

known

variant;

Claerhout, Afrikaanse Kunst, Utrecht 1971:

B Suguni region:
pi.

height

55cm;

C unspecified:

height

42 -Zcm. After A

54, 56, 57

Over the long term, whole art styles show the effect of major drifts, major
changes in taste. Coptic art began as a variant on Hellenistic art in Egypt,
originally being at the start close to an extreme naturalism. Gradually, the
icons were stylized to extremes. Themes and motifs of Hellenistic origins
remained, but in abbreviated stylized form. The idea was still there, but the
representation was no longer as full. Thus, a textile showing the head of the
goddess Earth in the beginning of the sequence with its intricate use of colour
for modelling and shading should be compared to late Coptic textiles where
women's faces, still held in a medallion, have become rectangular and suggest
the natural angles produced in woven stuffs more than the natural angles of a
human head. This progressive stylization is evident in all branches of Coptic
146

THE CREATIVE PROCESS

Fig. 8.4 The martyrdom of St Thecla. Limestone

9th

c.

St Thecla

is

relief.

Brooklyn Museum,

about to be burned and devoured by hounds. Coptic.

By

stylization is complete, a horror of the void (filled by flames and vegetation)

After K. Wessel 1965:59, pi. 62

New
this

is

York. Height 53

Scm;

length

period the relief is reduced

to

58 -Scm.

two planes,

apparent and the effect becomes decorative.

art, textiles, even painting. Taste changed

due to an effect of Christianity with its emphasis on the
inner life rather than on earthly pleasures? Who could prove it? The fact is
simply that major stylistic drift occurred (du Bourguet 1967; Wessel 1965;

art,

sculpture,

continually.

Badawy

Was

decorative
this

1978).

In several sequences of styles, one can follow the rise of decorative

elements until they swamp all empty space and even push back the figurative
parts of the icon. This is evident in the Benin series of heads and in the
treatment of walls in the madrasa of Morocco. Kuba art also shows a growing
tendency to avoid a horror vacui, i.e., to leave blank space. Such similarities in
development have been attributed to changes in taste stressing wealth, rather
than aesthetics and are commonly associated with centralized systems and
courts. This is certainly not a universal rule in art, but several parallels can be
found in or out of Africa, just as parallels for stylization in Coptic art can be
found. What such cases show is the inherent power of form and shape to
develop its own logic. The logic of geometrical decorative motifs is to expand;
the logic of an art that reaches extreme lifelike representation is to reverse the
trend. Parallel sequences merely show that the given styles logically only have a
few major possibilities for evolution. They can remain stable overall as, for
example, in the Faras sequence; they can move slightly back and forth from
their inherent norms as in ancient Egyptian arts, or they can move decidedly in
one of a very few directions such as towards or away from simplification.
Having given examples of simplification, we should consider a contrary case:
the goldweights in Ghana. They started out as purely simple geometric shapes
and styles. They ended in the nineteenth century at the extreme of complexity
147

ART HISTORY IN AFRICA

in line, volume and figure, often having become much more intricate than their
use would have warranted (Garrard 1980:274-315; see Plate 11.2). As in the
Coptic case, we can only guess the reason.

PRIME WORKS
Change does not always occur by sheer drift. Quite often a new type of icon was
suddenly created and became a new fountainhead for replication. Such works
are called prototypes or prime works. Such works constitute a clear break in a
tradition. The mosque at Qairawan provides a good example, involving only
moderate innovation and yet leading to a new prototype. Its plan derives from
the plan of the Madina mosque, its mihrab was not the first one to be built,
minarets of some form had existed before and it was not the first domed
mosque. But the overall shape of its minaret, the oldest now extant in Islam,
became the prototype of all later minarets in northwest Africa. Its innovation
in the creation of a central nave, the execution of its arches and beamed roofs,
the choice and execution of the main dome in front of the mihrab (Lezine
1966:62), were all perceived as so new and so perfect that these features were
not only imitated quite rapidly in the main mosque at Tunis, but further
inspired developments at Cordoba and in Morocco. Its influence lasted for

many

centuries.

radical. The appearance of round kidumu masks

among the Tsayi c. 1870 is such a case (Dupre 1968). The mask
and the dance for which it was made were invented then, representing perhaps
the turmoil of 'new times', but we do not know where its inspiration could stem
from. There was nothing like it anywhere in the area. Masks were known,

Sometimes the break was

(see Plate 7.2)

colours were used in other works and on masks, some of the decorative
patterns can be found on objects of a different tradition, but not the round flat
shape of the mask, not the strict double horizontal and vertical symmetry, not
the extreme abstraction of the face. Someone invented a prime work out of

nowhere. The apparition of such prime works in sculpture was usually almost
as radical as this, but in architecture or in Christian painting such radical
innovations have been much rarer.
The greater the break the more difficult it becomes to trace the sources of
inspiration. A typical case would be the explanation of the origin of the
ancestral statues north of the Sankuru in Zaire. They are attributed to the
Ndengese people and seem to descend from a nineteenth-century prototype,
the date being guessed at by the relative rarity of the extant pieces (Cornet
1976). The heads were very similar to those of the cephalomorph cups of their
southern neighbours in the settlements on the banks of the Sankuru and
further south. The statues have no legs but a broad base around the genitals.
That feature recalls the clay statues made by their northern neighbours
(Hulstaert 1931), as do the overall proportions of torso and head. The pattern
of scarification of the bodies and the caps were probably locally inspired. The
idea of ancestral statues for some great persons stemmed either from the north

Kuba kingdom to their south, where the royal statues of the Kuba
were created in the eighteenth century (Vansina 1978:212-15; Rosenwald
1974). There is enough evidence of social intercourse between the Ndengese,
their northern and southern neighbours, to allow for such influences.

or from the

148

THE CREATIVE PROCESS

But why did the Ndengese feel the need for such works? In the nineteenth
century they had no chiefs but were governed by an association of the wealthy,
the etoci. Yet the statues were not intended to represent all the etoci. There are
far too few works of art for that to be true. Was there then a further principle of

Plate 8.2 Siaiue. Wood. Use unknown, perhaps commemorative for leaders, perhaps for a
Dekese, Middle Lokenye, Zaire. Thought

to

be

commumty

cult.

Near

work of the Ndengese people. Komnklijk Museum voor Midden Afrika,

Tervuren. Height 139cm. Acquired 1912

149

ART HISTORY IN AFRICA

Wood and a bead in front of each ear. Used as a household cup for patnctans. Kuba
Volkerkunde, Berhn-Dahlem. Height23cm. Acquired 1906. The detail (keloid tattoo, top, eyes)

Plate 8.3 Cephalomorph cup.

kingdom.

Museum fur

favours an attribution

150

to the

northern

Kuba

near the Sankuru

THE CREATIVE PROCESS

Fig. 8.5 Shrine figures.

Mud. Wagania

village near Befale, Zaire. Height unknovm. 1913. The Nsongo people
shnnes at the entrance of the villages. The figures were placed on a hardened mud floor in a
shelter open to erne side. They were surrounded by a mud leopard and a mud dog. Shield, basket,
hat, cloth and a
bedframe against the back wall. Other peoples further south such as the Mbole north of the Ndengese made similar

(Mongo group)

figures. After

built such

a drawing by Gregoire in Revue Congolaise, 1913/14, 4: 181

Or was it a single workshop or a few workshops in adjacent villages

produced them? Fieldwork can perhaps still yield clues to answer these

selection?
that

questions.

How

did such prime works come into being? As commissions? Very

had never been seen before. One suspects that the creation

unlikely, since they

of masterworks was responsible. Carvers among the Kuba used to make

extraordinary icons as bravura pieces, to attract fame and business, or merely
out of fun, or for some pious reason, as in the case of the man who carved a
statue of his pregnant daughter, a type of icon hitherto unknown (see Plate
6.6). Cups in the shape of equestrian figures astride antelopes or a dish
supported by the lower half of a female torso with three persons kneeling on its
rim are obvious bravura pieces, destined to make the sculptor known. The
antelope horn in wood which imitated the spiralling line neither of one type of
antelope, nor of another but a line in between was a piece made for fun by a
carver who would exhibit it and ask which antelope's horn it represented! Such
isolated bravura type works can be found almost everywhere. So perhaps they
were the seminal inspiration for prime works. If so, the early attempts to create
prototypes were destroyed by the carvers until they were satisfied with the new
prime work.
But prime works became prime works only if they were copied. Why

151

ART HISTORY

IN AFRICA

Plate 8.4 Statue of king Shyaam aMbul aNgoong. Wood. Used as a commemorative statue on display in the palace.
Also possibly as a charm to facilitate childbirth in the royal harem and as a receptacle for the breath of the dying king.
British Museum, c. 1750. This is one of the earliest carvings in a set of ndop, as they are called

152

THE CREATIVE PROCESS

Plate 8.5 Vessel. Wood. Used as a household object for patricians.

18cm. Acquired
that part of the

in

nm

Africa 1906. Northern

Kuba

and one figure have been

three figures nor, if so,

Kuba kingdom. Yale

near Sankuru. This

'totally restored' , i.e.

whether the third one was similar

to the

two

is

University

An Gallery.

Height

the vessel in Frobenius 1907: pi. p. 233.

Note

added.

We

others.

Meaning, probably not symbolic. Possibly a

are not certain that there

were originally

pun

153

ART HISTORY IN AFRICA

would masterpieces be adopted? Often the acceptance of a new prototype was

new institution or to the fact that it perfectly

exemplified a concept relating to a major institution. Thus, the rise of Kuba
royal statues is to be linked with the decline of regional cults and the rise of a
cult for the king as 'nature spirit'. The statues were a step in this evolution. We
will never know if a king thought of them and commissioned them, or if a
bravura piece brought the idea into a king's head. The edan bronzes of the
Ogboni cult, the single most important organ of government in many cities of
western Nigeria, must be linked to the very development of the cult (Drewal
1980; Morton Williams 1960; Williams 1964). The intended use of any class of
icons is therefore the only clue we have as to why masterpieces could become
linked to the appearance of a

prime works.
Drift and sudden mutation both occur in the visual arts. To a certain
degree, every icon is both a replica and an innovation, never wholly original,
rarely a totally slavish copy. Comparisons determine what are unique items:
masterpieces never imitated, or prime works, and what are replications. The
ability to find the relationships between different works in such terms is crucial
to art history. One must assess not only innovation, but the magnitude of
change as well. For change in art is bound up with other cultural and
concomitant social changes. Not only the nature, but also the magnitude of
change in art can be clues to the nature and magnitude of concomitant change
in society and culture as well.

CREATIVITY, SOCIAL

AND CULTURAL CHANGE

Society, culture and the arts are in reality so closely intertwined that it seems
evident that change in one must be accompanied by change in the other. Art is
an integral part of culture and individuals participate in social life through the
medium of culture (Layton 1981:187). Moreover, visual art is often directly

So the axiom
seems well founded.
A good example of these relationships is the case of the Fatimid portals
discussed in Chapter 5, pages 97-9 (Sourdel-Thomine and Spuler 1973;
Hrbek 1977). After the conquest of Tunisia in 909, the Fatimid ruler did not
settle in pre-existing capitals for long. He built his own city, Mahdiya, so called
because he claimed to be the Mahdi, the Messiah who announced the end of the
times. As the leader of the Shi'i branch of Islam, he was considered to be a
divinely inspired, almost God-like leader (/mam), a sacred king. In this city he
built his mosque and the first portal appears in front of the mosque (A.D. 916).
The jx)rtal may well have been inspired by Roman triimiphal arches nearby as
Lezine (1966) argues. That would be fitting for the almost-divine king. The
founding of a new capital itself was not an innovation. Early Abassid Caliphs
and their representatives in Tunisia, the Aghlabids, had done the same. When
the Fatimids conquered Egypt (969), they founded a walled town nearby other
such foundations and called it al-Qahira, 'The victorious', which became
Cairo. There, too, they built congregational mosques and the portal reappears.
The portal became a constituent part of any congregational mosque officially
commissioned by a ruler until well after the fall of the Fatimid caliphate, the
last known one being built by Baibars, the first Mamluk Sultan in 1267-69.
tied to specific institutions,

which use

art objects as their tools.

that change in art reflects sociocultural change

154

THE CREATINE PROCESS

Plate 8.6

Main

portal.

Sultan Hassan Madrasa and

vault. Stone masonry. Cairo.

Height 34 -Sm. The arch

Tomb Complex. Small semidome with stalactiu (muqarnas)

20m high. 1361. Shape ofportal introduced by Baibarsfor his

is

madrasa 126213

155

ART HISTORY IN AFRICA

Later rulers no longer commissioned congregational mosques and the portal
does not reapjjear. It has been suggested that this was due to the lack of space in
or nearby the city within reach of a congregation. But other reasons are more
compelling. For Mamluk sultans greatly favoured large complexes of tombs
and madrasa, which had been introduced in Egypt just after the ousting of the
Fatimids by Saladin (1171). As a good Sunni - the name designates the other
major branch of Islam - he fought the Shi'i heresy by introducing the madrasa,
schools of higher learning where theology, law and the Arabic disciplines were
taught. The madrasa complex in architecture had first been developed in Iraq
to house the institution. Coming from Syria, Saladin knew about it. Soon
domed tombs for royal patrons were attached to the complex. The Mamluk
domed tombs of Cairo derive architecturally from the Turkish regions whence
they came and the name for such a tomb, turbe, is Turkish. The whole complex
then is not a mere reaction against Shi'i doctrine, but reflects the cultural
origins of the Mamluk.
The madrasa complexes came to develop a portal of their own from c. 1262/3
onwards, the most famous being the ponal of Sultan Hasan's complex
(finished 1361). It greatly differs from the Fatimid model. Madrasa portals
derive from facades in the Konya area of Turkey and, like the tombs, reflect
the cultural origins of the Mamluk rulers. But both Fatimid and Mamluk

major religious buildings, commissioned by the

portal was the successor to the Fatimid
structure and the people of Cairo were certainly aware of the continuity, which
left a trace in the structure. The Mamluk portals seem deeper than their
prototypes in Turkey (Sourdel-Thomine and Spuler 1973; Hrbek
1977:10-69). From these brief indications, the reader can see the interplay
between large-scale soccial change, large-scale change in culture (religious
ideas), and the evolution of portals. The art form followed socio-cultural
change and expressed it. In this case, art was an epiphenomenon.
But does change in art always reflect prior change in culture and society?
It can be argued that drift is autonomous. Layton (1981:178) lists among his
sources of creativity the act required to give an idea tangible form - even a copy
is not identical to its model. He also mentions elements of design that are not
culturally significant and where free variation is allowed as well as the
expression of fresh metaphors. It might be accepted that the appearance of new
prime works, i.e., masterworks that are adopted by the public and therefore
copied, are different from cases of drift. The latter could be autonomous
developments. Thus, when the Kuba adopt a variant on an earlier pattern and
recognize it as a new design, by giving it a name, this need not reflect any
change elsewhere. Nevertheless, even here the practice is related to the
encouragement of innovation in design, itself related to the growth of the
patrician class, its need for novelty to offset itself from others and to fuel rivalry
between the patricians themselves, and, in turn, this is related to the growth of
the Kuba bureaucracy from the eighteenth century onwards (Vansina 1978).
The invention is autonomous but it reflects a climate of historical change. The
changes in the style of the Faras wall paintings, which we discussed in Chapter
5, page 95, clearly reflect wider - and prior - cultural change. They are
portals

were

built in front of

ruler. In that sense the

Mamluk

linked to the variable influence of Coptic monasteries as against the ordinary

hierarchy of the bishopric, the poUtical and ecclesiastical variations in

156

THE CREATIVE PROCESS

relationships with

Muslim and Coptic Egypt,

access to Jerusalem and Syria for

pilgrimages, and contacts with Byzantium. Stylistic change, moreover, was

accompanied by changes in the languages used at Faras (Greek, Coptic,
Nubian and Greek/Coptic alphabets), and the appearance of new themes,
mostly the depiction of new saints. Coptic saints appear in periods of Coptic

and Byzantine saints

became important (Michalowski 1967). Arguments
have been developed around the graphic signs on the paintings that allow us to
deduce changing over from Monophysite (Christ has one nature) to Melchite
(Christ has two natures) beliefs. These are detected in the varying depictions of
the madonna. No doubt changes in styles at Faras also reflect wider and prior
influence, Syrian saints in periods of Syrian influence,

when Byzantine

influence

changes in culture and society.

The adoption of new themes or new prime works means the adoption of
new icons, new metaphors, which indicate in turn either the creation or the
development of new ideas. Very often this points to institutional development,
even
as most works of art are linked to institutions. In this regard,

masquerades are institutions, and the appearance around 1870 of the round
Tsayi mask (Plate 7.2) reflects social change. The masquerade was then
created to express the insecurity of the times. Before that time, the Tsayi had
been prosperous, mainly in the slave and ivory trade. From that time onwards,
they began to be bypassed and beset with competitors, not only for their trade,
but also for their land (Dupre 1968).
The autonomy of art lies in its formal evolution. It was perhaps a necessity
for Kuba kings in the eighteenth century to develop royal statues to
commemorate former kings, and to express an evolution in convictions by
which kings were now seen not merely as priests of nature spirits, but as

Hence the reason for the adoption of the

statue per reign (Rosenwald
commemorative
practice to commission one
of the ruler, was an
representation
of
form
the
itself,
of
art
work
the
1974). But
autonomous invention, deriving from earlier forms such as tiny statuettes used

powerful nature

spirits

themselves.

territory near the border to represent nature spirits

and by larger statues that probably existed - as they do now - in component
do not know exactly
chiefdoms to honour an ancestress of a chiefly line.

in a portion of the

Kuba

We

what the sources were

were local works.

for the

new

icon, but

we can be almost

certain that they

Even Kuba bravura pieces, which best exemplify the autonomy of formal
creation, are still Unked to a climate where innovation is prized. They embody
the competition between carvers and the dynamics of competition between the
patricians. They bought such works as a sign of prestige or power, and they
encouraged artists to produce new icons, to be used as counters in their own

competition for the display of influence. That could only happen in a kingdom
whose bureaucracy was constantly expanding.
Lastly, one may wonder whether the visual arts themselves ever led to
change. Is art always a passive epiphenomenon? As new metaphors were
created,

new

tools

became

available for social or cultural use.

For instance,

every development in body art, textiles, costume or household goods, could be

used to stress differences between social strata, and innovation was encouraged
when the urge was felt to express such increased differences. Art works as
concrete symbols could crystallize unfocused ideas and mobilize people. There
157

ART HISTORY IN AFRICA

Perhaps the most
Anthony in the former kingdom
of Kongo, where the common people adopted them as a symbol of their
understanding of Christianity as a truly African and not European church, in
are very few records of such an evolution, however.

sensational

is

the use of existing statues of St

turn linked to their opposition against the ruling class and against the
lawlessness and insecurity produced by internecine conflicts led by this class.
St Anthony had previously been a favourite saint of the European missionaries
and had been especially invoked by the Portuguese for lost causes and for the
downtrodden. He now became the most important saint in heaven; he was seen
in the statues as a black person carrying the child Jesus (Randies 1968: 149-51,
157-60; Balandier 1968:263). But even this example is not conclusive. The
new conception was a development of the old, and it did not occur before the
ideology of 'Antonionism' developed, but concurrently with the larger cultural
reinterpretation. It is a good example because it shows an interaction between
art and culture that is the answer to our question. Art as a crystallizer of
in tangible form could lead to change by focusing ideas and by
mobilizing support. But art could only do this as part of a wider cultural
system, as a reaction to some action outside, which art then in turn influenced.
Art is an epiphenomenon but epiphenomena can sometimes take the lead by

metaphors

the effects they produce.

158

CHAPTER NINE

THE CREATIVE PROCESS:

FOREIGN INPUTS
DIFFUSION: THE

MEANS

Creative processes can be an internal process only (Kroeber 1948, 1953; Sapir
1916; Graebner 1911; Vansina 1 96 8) , as we saw in the previous chapter , or they

can be induced, at least in part, by external stimuli. Foreign objects,

techniques, stylistic characteristics, themes or motifs could influence the taste
of the public and set artists thinking. Quite often such influences enriched the
local repertoire, and in art as in other fields they were quite important. If the
oikoumene had such a varied output, it was largely due to the fact that its artists
had access to a larger storehouse of forms and concepts from which to choose.
The notion of regional tradition itself implies that mutual exchange between
neighbouring styles did not reach very far and isolation prevailed. Hence,
considerable originality was preserved, but a major voice for inspiration,
external stimuli, remained mute.
Historians have shown over the last generation that the much-touted
isolation of African communities was largely illusory. There were always
relations between communities and societies, stretching like a web over most
of the continent and of a more intensive and longer lasting nature than is
realized even today. If this is so, mutual borrowing must have played a much
more substantial role in art than has appeared hitherto. On the one hand
traditions might be less local than they seemed to be at first sight; on the other,
the persistence of great originality of expression in neighbouring and
communicating communities needs to be explained.
Borrowing of foreign objects certainly occurred. A ewer of brass of
Richard II of England (1377-99) was found in coastal Ghana, no doubt having
arrived there by the caravan trade from North Africa (see Shinnie 1965: ill. 88;
Fagg 1970:51). Foreign objects were also copied. Two copies of Coptic lamps
have been found near the northern fringe of the Akan speaking area in Ghana
(Arkell 1950). The originals, probably made between A.D. 300 and 700 also no
doubt found their way by caravan trade. Chinese celadon abounded at one time
on the East African coast, and there are reports of the export of textiles from
Benin on the Loango coast (Martin 1972:63). Knives from the Ubangi bend
area (Fig. 9.1) were used as insignia just north of the Malebo Pool more than
500km downstream. Such cases can be multiplied. Foreign objects were both
imported and copied.
The process of borrowing involves first the means to acquire foreign
stimuli, usually objects, and secondly the demand for their imitation or new
features inspired by them, a demand that was always internal. Trade was the
most powerful mechanism for spreading foreign objects. Intercontinental

159

ART HISTORY IN AFRICA

copper and brass. Ngabe, Congo. Height 49-5cm; width ISScm. Drawing
964. The Tio near Brazzaville acquired such knives of honour from the Ubangi nver as a byproduct of trade. The

Fig. 9.1 Ubangt kmfe. Iron, handle

field

original symbolism (female sex)

is lost

transit

is responsible for the wide diffusion of cowrie shells from

the Indian ocean over very large parts of the continent. Already in Roman
times the shells were traded from the Maldive Islands to North Africa. Later
they became the currency in interior West Africa and seeped inland in portions

trade, for instance,

of eastern Central Africa. It was only in southernmost Africa that the cowrie
did not penetrate. Trade was the artery binding the countries of the
intercommunicating zone, the oikoumene, together. Between this region and
other parts of Africa, important channels of trade existed from early times. In
the Indian ocean, trade between East Africa and the Red Sea is reported by the
first centuries A.D. Trade from the Sahara and beyond created the wealth of

Leptis Magna, on the coast of Libya, during the same centuries. After the Arab
conquest of northern Africa, a trans-Saharan trade, fuelled by the demand for
gold, developed on a large scale and ultimately brought the whole of West
Africa into contact with the oikoumene. In East Africa intercontinental trade
was first limited to the coast but ultimately, as interior trading networks
developed, bye. A.D. 900 affected central Africa (Zambia and Shaba) as well as

Zimbabwe and even

the

Limpopo

area. In Equatorial Africa the interior

network of trade along the Zaire grew under the stimulus of the expanding
slave trade after

c.

1530.

for the dissemination of foreign objects were also

In northern Africa pilgrimages were very important: Christian
pilgrimages to Jerusalem, as well as Muslim hajj to Mecca, not to mention

But other means

common.

minor but favourite

sites

of pilgrimage such as the shrine of St

Menas

(before

from where typical bottles

Tunisia and even to Europe (du Bourguet

A.D. 400), Egypt's national saint, west of Alexandria

were exported

to as far

away

as

1967:81, 90). Pilgrimage explains why the battle standard in seventeenthcentury Ethiopia was an icon of Christ with the thorn crown, brought from
Jerusalem (Doresse 1972:15, 85, 117, 216). Legend has a ruler of Mah
bringing an Andalusian architect back to the Niger from his hajj. South of the
Sahara, however, pilgrimages, where they existed, were much more localized.
Other religious concerns could lead to dissemination here. Thus, in the 1870s,
a Loango healer visited the area near Cape Lopez and carved statues in his own
style to serve as protective charms. Later they were moved to near the Gabon
estuary (Nassau 1904:308-11). In 1902 a Kuba king commissioned a similar

160

THE CREATIVE PROCESS: FOREIGN INPUTS

charm from a renowned medicine man of eastern Kasai (Songye) and his statue
was displayed for years at the very hub of the Kuba capital (Vansina
1969b: 19-20).
Besides trade and religion, occasionally other mechanisms were
responsible for the diffusion of alien objects. Travelling entertainers ranked
high among these at certain periods, and introduced both dances and masks to
places that had not known them, with or without cults. Olbrechts (1959:26)
noted such dynamics as early as the 1950s. Another of these occasional
mechanisms was state expansion, and the concomitant spread of its dominant
is clearly responsible for the spread of so-called Luba
and especially the caryatid thrones, as the Luba
or
motifs
styles, themes
empire grew in the eighteenth century. An oddity in the same area is the case of

culture. In Shaba, this

Lunda empire where certain objects such as swords spread as it expanded,

but where no Lunda art developed at the centre (Crine-Mavar 1968; Neyt
1981:219). Rather, the dominant group relied on the import of art from their
western reaches, from central Angola. The larger empires elsewhere in Africa
must also have acted as disseminators of art. In the case of Benin, this accounts
for the occasional use of Benin objects in shrines all over the Niger Delta
beyond Benin's realm proper (Horton 1965b; Alagoa 1972: index 'bronze' map
186). In southeast Africa, Zimbabwe-type buildings in the provinces of the
country are witnesses to the same mechanisms (Garlake 1973).
One mechanism which was once often invoked must be rejected: the
effect of migrations, which were believed to have transplanted whole material
cultures in a short time over huge distances. Such spectacular changes have
been very few in Africa. Migrations may have killed more artistic traditions
than they inspired novel ones. Thus Sape sculptural traditions were ended by
the Mane invasion in Sierra Leone (Page 1977:508-9). The clearest case where
immigrants implanted a truly novel material culture was in Madagascar.

the

DEMAND FOR FOREIGN WORKS OF ART

foreign objects were in a community, and there must have been some in
African communities, they could be imitated but that would occur only if
there were a demand for them. Typically, some were rare prestige items,
linked to ruling groups in states or to big-man status elsewhere, for example
the shells of the Indian Ocean conus that were rare, costly and visible signs of
prestige in the interlacustrine area and as far west as Lake Mayi Ndombe in
Zaire. Foreign objects, such as a 1785 ship's bell in the nineteenth-century
treasure of a Gabonese king, or European crowns in the hands of big men on
the Gabon River, and the monarch of the Ogowe Delta, are as typical as the use
of a Benin mask by the nineteenth-century ruler of Igala on the lower Niger (H.
Bucher, personal communication; Ben Amos 1980a:20-l). The equivalent of

Once
all

European curiosity cabinets existed in some African courts as among the

Kuba.
Such prestige items, where possible, could be copied. The case of the
double bell, an emblem linked to political and military authority, is striking.
The item was first made along the West African coast somewhere between the
Ivory Coast (its probable point of origin) and Benin. But it spread across the
forest all the way to great Zimbabwe, where it was in use before A.D. 1450. It
161

ART HISTORY

IN AFRICA

Key
Single

^^^

+ double

bells

Single bells

Fig. 9.2 Distribution of double bells

Central Afrtca

was imitated by local smiths everywhere as far south as the Zambezi, but
Zimbabwe imported its bells apparently north from the river. In the Ubangi
bend, the production became large enough so that double hells began to
function as a currency (Vansina 1969a and Fig. 9.2).
Local workshops copied a whole item or a technique, motifs or a theme
according to local demand. Thus, lamps in ceramics from North Africa,
probably from Kabylia first, reached northern Nigeria with the trade, and
local populations as far south as the Middle Benue adopted such lamps. They
did not exactly copy the North African models, but it is still easy to recognize
the kinship between a Hausa lamp or one of the Middle Benue (Jukun, Ibi) and
its North African prototypes, by general appearance, perhaps by the use of the
colours common in North African lamps but also common in Nigeria, but
especially by some of their shapes and motifs (Krieger 1965-69: vol. 2; ill.
207-9, 219-20). In this case, no emblems were desired, but a new utensil was
widely spread. Nevertheless, it is more than likely that it first spread among
the dominant groups.
Goldweights had to be borrowed by Akan traders in gold, if they were not
to be cheated. They borrowed from the middle Niger and borrowed shapes and
motifs with them. Later, however, they rendered European objects and
162

THE CREATIVE PROCESS: FOREIGN INPUTS

motifs in brass goldweights and in gold ornaments as well. Thus some weights
occur in the shape of Dutch teapots. Indeed, the Akan lion developed out of

English lions (Ross 1977).

Copies were rarely perfect and did not need to be; if utilitarian objects
were practical, they would do. Prestige objects or objects for entertainment
would conform to local taste except that the prized foreign features would be
retained, as in the case of the Nigerian lamps or of arabesque forms of
decoration in West Africa. Here the decoration, as on the nineteenth-century
walls of Akan palaces is recognizably inspired by North Africa, but
nevertheless quite different (Denyer 1978:79-81; Prussin 1980: esp. 71).
Demand for utilitarian objects and for prestige items was supplemented in
some cases by religious demand as well. Converts needed mosques or churches
built them after the pattern of those who converted them. Thus, Schacht
(1954) was able to show Ibadi Muslim influence in West Africa, because Fulani
mosques there lacked a minbar, the staircase/pulpit, a typical Ibadi feature,
even though today Fulani are Sunni Muslims and have been so for centuries.
The importance of a demand is highlighted when the existence of factors
blocking diffusion, despite extensive contacts, is recognized. Diffusion and
borrowing are not automatic processes at all. The presence of Catholic art
forms in Mozambique and even in Zimbabwe during the seventeenth century

and

did not lead to a diffusion of Christian iconography nor even of Christian

principles of sculpture. In lower Zaire, the Christian art of the upper classes
did not influence the popular arts, but the reverse occurred probably after 1665
and the collapse of the old kingdom. The relationship between Islamic art and
other arts in West Africa, where there has been intensive contact for nearly a
millennium, is even more striking. Apart from decorative art and some
architectural designs, Islamic artistic influences on the main arts south of the
frontier have been slight. In turn. West African sculpture did not
influence northern Africa at all. Even in architecture it is much easier to see
that the great city gates from northern Nigeria were not copied further south
than to fmd positive links between architectural plans. Those that are found in
western Nigeria may be more closely compared to the classical impluvia houses

Muslim

Africa (Ojo)! The main reason is that Muslim tradition

and that western African peoples prized representational
sculpture and used 'idols' profusely in their religions. West African sculpture
and Islam could coexist, as Bravmann (1974) has shown for the middle bek of
Nigeria (Nupe) and for northern Ghana, and as can be seen from toys near
Lake Chad and in the Middle Niger belt. Nevertheless, as Islam gained,

of northern

condemns

Roman

idols,

sculptural traditions receded,

probably because the traditional religions

receded.

Moreover, Islamic architecture in West Africa did not slavishly follow

Northern African Saharan practices. Wherever, as in the Middle Niger belt, an
autochtonous technique (mud architecture) and tradition of architecture had
existed before the arrival of Islam (Gruner 1977; Prussin 1968), Muslim
features were translated into the traditional medium and even traditional
shapes were often conserved. There was no perceived need literally to copy all
the features of North African building. In reverse. West African dome
construction and mud architecture may conceivably have influenced northern
African styles, either in Roman times as seen by Roman concrete construction

163

ART HISTORY

IN AFRICA

or even as late as the ninth and tenth centuries,

filletted

when

it is

conceivable that the

used in Tunisia, was in fact borrowed from the Sahel. This

has been ascribed Iranian origins but the chronology is unsound

dome,

first

type of dome
(Lezine 1966:79-88). No known West African architectural influences
travelled north later. Not only were the inhabitants of northern Africa
Muslims, but they felt so superior to West Africans in other respects, that they
had no desire to copy anything, except in the realm of music and dance.
A similar barrier is evident in Ethiopia. Muslim and Christian antagonism
here minimized artistic interchange. It occurred but was restricted to minor
features such as the decorative margins of manuscripts or embellishments in
buildings. Christians copied arabesques, hareg (Leroy 1967, 1968, 1973;
Chojnacki 1973) and eleventh-century Fatimid embellishments found their
way around some windows of Lalibela churches, where keel arches, beloved by
Fatimids, are occasionally found as well. On the whole, the refusal to borrow,

however, is the striking conclusion.

In most instances one should not attribute lack of borrowing between
different regional traditions to lack of contact, but rather to active cultural often religious - barriers. Conversely, where borrowing occurred, its
explanation requires a consideration of the reasons why it occurred. Just an
inherent belief in automatic diffusion will not do.

DISTRIBUTIONS*
Claims for borrowing are arguments based on the presence of distributions that
exceed the range in which the products of a workshop or a set of workshops are
used. In the past, distributions have been established mainly when it was
suspected that the range or a feature of an icon exceeded the range of an ethnic
group. But ranges coextensive with the claimed extent of an ethnic group
should not be postulated as freely as they usually are. The situations are much
more dynamic than that. Thus, in the Kuba kingdom, the distribution of
cephalomorph pipes belongs to the central and eastern parts and is found
outside the kingdom in the east. The use of double-headed cups is restricted to
the west of the kingdom and beyond its borders between Loange and Kasai. It
occurs far to the south in another 'area' that of the Pende (Himmelheber
1960:733, 383-4; De Sousberghe 1959:15, 140; see Fig. 9.3). Ideally, the unit
to examine is the workshop or the village only. Lack of data forces us, in
practice, to consider whole ethnic groups.
The first point to examine must be the criteria by which distributions are
plotted (Vansina 1968). These are always selective. They include features of
style, theme and motif, usually in relationship to whole works of art. Use alone
is not a valid criterion because it does not necessarily relate to a given object,
thus, 'emblems of leadership' are a meaningless category for comparison. Use
criteria is valid, such as 'adzes as emblems of
But the main relevance of use is to show that certain features of
the objects compared are or are not arbitrary. Slits to see through in masks

tied

to

some formal

chieftainship'.

destined to be worn are not arbitrary. They are essential for the wearer of the
mask. But, say, an elliptic shape of the mask is arbitrary and hence valid for

comparison.
*

See Schmitz 1967.

164

THE CREATIVE PROCESS: FOREIGN INPUTS

K U B

Fig. 9.3 Topogram of the distribution of doubleheaded cups and cephalomorph pipes in and around the Kuba kingdom.
'Kuba' can designate the kingdom or a wider area. 'Kuba' objects are not homogeneously distributed over either area

if the features compared are

and the more arbitrary they are, the stronger the proof. The
fundamental criterion for a valid distribution is the criterion of quality. It
refers to several independent arbitrary features in the artefacts compared. One
such feature may be due to chance, just as a concept may by chance be rendered
by the same word in two entirely unrelated languages, but several common
features cannot be attributed to chance. An example commonly given is that of

Distributions will only reflect borrowing,

arbitrary,

the

game of chess. The number

of squares used, the types of pieces used, the

by which they move, are all arbitrary. To find the game in Iran and in
Europe cannot be accidental (Kroeber 1948). Similarly, if all the
characteristics of two churches correspond: from orientation, to cruciform
shape, to presence of an apse in the same position on the plan, to the details of
clerestory windows, and so on, then the churches must be related. If two pieces
rules

of sculpture correspond in general position in space, in proportion, in the

carving of major detail and in small detail, especially 'irrelevant' detail, they
must correspond. Here the major criterion for establishing affinity is the one

used in morphological analysis and to establish stylistic series.

Another criterion, the criterion of quantity, is useful after the criterion of
quality has been established, but it can never be used alone. This criterion
states that if several different artefacts, or techniques, or non-material features
are found, each individually quite similar through arbitrary features to
artefacts, techniques or non-material features in another locale, then all these

165

ART HISTORY IN AFRICA

objects, techniques or intangible features are related. In fact, the criterion
either strengthens the case for borrowing for a single type of object, or builds a
circimistantial case for intensive connections

objects were found.

between the places where the

An example of the first situation is the comparison of Kuba

velvet cloth of the nineteenth century with velvet cloth

from the Zaire coast

dating from the seventeenth century. The patterns correspond (cf. Plate 9.1
and Fig. 7.3), the techniques correspond, and the term to designate such
objects corresponds (Vansina 1978:220). The linguistic evidence, quite
independent, yet tied to the object, is decisive. An example of the second
situation is the town of Begho (Nsoko) in Ghana when compared to the Upper
later. Not only were similar shards of pottery found,
but similar ceramic goldweights occur, foreign objects from North Africa
occur near Begho, techniques for weaving cotton were apparently similar, and
filigree techniques from the Niger were imitated at Begho. The evidence adds
up to an impressive set of similarities in unrelated items, not easily invented
independently. It points to intensive contacts, and these indeed existed from
perhaps 1350 onwards. Gold from Ghana moved through Begho to Jenne on
the middle Niger and did so until late in the nineteenth century (Posnansky

Middle Niger c. 1400 and

1979).
it is quite difficult to establish that borrowing must explain the
independent invention could be equally likely. A widespread
decorative design known as the Hausa knot in West Africa consists of a plaited

Sometimes

distribution as

Plate 9.1 Cloth. Raffia. Pile work, with dyed thread. Used as cloth for aristocracy. Lower Zaire, former kingdom
Kongo. Ulmer Museum. 219 ^ I7Scm. Before 1659. The patterns recall those current c. 1900 in Mayomhe (northern
Kongo) as well as some on ivory horns C.ISS2. Technique as in modem Kuba work, which derives from this type of work.

Kuba

166

patterns are also geometric but not identical with these.

THE CREATIVE PROCESS: FOREIGN INPUTS

of guilloche patterns often on an angular background. These are known in
northern Nigeria, Somalia, the Kuba country (imbol design) and North Africa,
from Byzantine mosaics onwards (Fig. 9.4). Even though the design looks
complicated, in fact it is not. It also occurs on early Iroquois pottery. Yet along
set

Fig. 9.4 Interlace knots.

Moroccan, miniature

C applujue and embroidery on

in book,

19th

c;

Akan

embroidery with muslitn magic square,

For a northern Yoruba instance see Plate 2.3.

These knots in West Africa probably stem from North Africa, whence they came as decorative patterns on textiles. In
North Africa they derive from such knots on mosaics and pottery since Byzantine and Roman times. Elsewhere in Africa
(East, Central) and Asia (India, Tibet, Indonesia) the motif may be independent - it is a simple pattern - although
Muslim influence in some areas also may be relevant. In North America (Iroquois pottery) the motif was probably
independent. A afterG. Vidalenc, I'Ari Marocain,Pans 1925: pi. XIV. B after Picton and Mack 1979:156 (acquired
1951). C after R. Sieber, African Textiles and Decorative Arts, New York 1972:38 (acquired 1948)

20th c;

mar^s robe, Kano, Nigeria, 20th

c.

167

ART HISTORY

IN AFRICA

with Other similar motifs it is hard not to consider that the continuous
distribution throughout northern Africa into Somalia on the one hand and
Nigeria on the other is just due to chance. In such a case the means for
borrowing must be carefully considered. Trade and especially the presence of
these motifs on clothes and leather belongings makes it quite likely that indeed
all the patterns from the Mediterranean to the Gulf of Guinea are related, but
not the isolated Kuba instance. There - until further notice - we must accept
independent invention. Guilloche was quite common in Kuba decoration,
whether in angular or rounded form and so was plaited guilloche. It was easy
for a

Kuba inventor to consider turning the design into a self-contained unit in

When one knows that the Kuba topologically did in fact invent most of

space.

the possible geometric combinations of patterns, the explanation becomes

convincing.
The arbitrary character, the possibility of independent invention and, in
an accessory way, the presence of unrelated similarities in other objects, are the
main criteria by which distributions should be evaluated and hence the main
criteria

by which they should be established.

RECONSTRUCTING DIFFUSION
a distribution indicates the likelihood of borrowing, the researcher must
then establish its date, direction, and mechanisms. He must provide a
hypothesis that will adequately and economically describe the borrowing
process. To provide all these elements is often quite difficult, and distributions
themselves must normally be complemented by other data. Thus, direction of
borrowing can sometimes be found by a consideration of related distributions.
When I am considering one particular type of throwing knife, very similar in
two places, one below the equatorial African forest and one north of it, and
even strengthen my case by showing similarity in the terminology, I still do not
know in which direction the knife spread. But in considering that it is but one
of many similar shapes in the north and the only one in the south, one can argue
that the feature was older in the north. It has had time there to agree more with

Once

have flourished to the point that all

direction then would be north
these other forms of throwing knives exist.
But the argument is easUy
Cordell
1973).
(McNaughton
1970,
to south
reversed. Perhaps the item travelled south to north, but acquired a new
significance in the north which led to an efflorescence of derived forms. Given
a respectable age, this could have happened. In this case we can be reasonably
local style in other matters

and especially

to

The

movement because the distribution ultimately

includes the throwing sticks seen on ancient Egyptian paintings and similar
objects on rock art. The distributions of various forms fall into two general
groups, an f-shaped knife and others, and the total distributions involved most
of the forest people (who cannot throw such knives!) as well as some groups on
its southern fringe. North of it the distribution was quite solid. A consideration
certain of the north to south

of the use of the weapon (in open land, especially in short grass steppes)
together with the archaeological data, makes it certain that the spread was old
and from north to south. The weapon may be Egyptian in origin or Nubian or
Saharan and it was first made in wood. When it diffused into the forest, the
data on use make it clear that it became a prestige item, linked to big men and

168

THE CREATIVE PROCESS: FOREIGN INPUTS

Key

ART HISTORY IN AFRICA

even south of the forest

it

did not turn again into a widespread

weapon

(Fig.

9.5).

The circumstances of use, the 'core' quality of the object in a society and
cuhure, the shape of the distribution, the shape of related distributions and
archaeological data all have some bearing on the previous illustration.

ASSESSING THE EVIDENCE

A few recipes for dealing with distributions as evidence, especially for dating
and origins

exist (Wissler 1926;

Hodgen

large distribution, well integrated, can

1942; Kroeber 1953). Thus a very

be a sign of great age. Broken

were never unified, must result from independent

two related variants where variant A either
encircles variant B completely or where the distribution of A is broken by B
into two or more parts, indicates that A is older. Overlapping patterns of
several related variants can be ordered in time sequence by treating them as
successive breaks of once-unbroken distributions, especially when other
criteria show that the tentative sequence forms a stylistic or technological
series. Distributions in the shape of a thread or as a sequence of spots in linear
pattern show by which route it took place and often give a clue as to the
carriers, such as traders, fishermen, pilgrims and so on (Fig. 9.6).
But none of these arguments, or less often used arguments taken from
that

distributions,

invention.

The

distributions,

is

distribution of

foolproof. Situations are

known

in general anthropology,

and archaeology where similar shapes of distribution occur and yet

have been shown not always to indicate the correct older, younger variants,
sequences of variants, origins and dispersal routes. On the whole, such
linguistics

indications are often correct, but not always. They can be used to elaborate a
hypothesis, but need to be backed-up by other evidence, either direct or drawn
from linguistic evidence about the names used for the objects in the cultures

and

societies involved.

hypothesis of diffusion must ultimately be tested by an attempt at

by assessing the likelihood of contrary or variant hypotheses. If a
particular explanation is only as likely as another or is only the most plausible
one of three or more likely possibilities, it must be discarded. It follows from
this rule that the more facets one considers in a reconstruction of borrowing,

Any

falsification

it becomes to arrive at
Hence, the hypothesis that fits all
the data becomes much more convincing, and it becomes all the more

the

more elements enter into play and

the

more difficult

different hypotheses that are equally valid.

convincing the more simple it remains. Elegance in hypotheses is not only a

logical virtue, but carries conviction.
Conversely, it is easy to show that proposed borrowings are false or
unproven if any improbability in dating, distance, mechanism of transmission,
circumstance of use, or reason for borrowing appears. Thus Baumann
(1969:57) sought parallels between statues from eastern Shaba (Zaire: Hemba)
and Sabaean statuary. The time difference of almost two thousand years, the
lack of direct communication between the east coast and eastern Shaba before
the nineteenth century, the lack of comparable objects in the whole
intervening area, as well as the great length of the indirect route via the

Zambezi that could have been used after c. A.D. 1000,

170

rule the

comparison out.

THE CREATIVE PROCESS: FOREIGN INPUTS

Moreover, it does not give any reason for borrowing and is based on too few
points of genuine resemblance.
In this it differs greatly from the imitation Coptic lamps in the shape of a
since
bird found in Ghana at Tarkwa. True, the distance to Egypt is great, but

m ^
Distribution of variant

Complex

distribution (languages)

6
Spread by fisherman

Spread by traders

Key
Oldest

River

case 'complex distributums' shows the

Fig. 9.6 Dtstnbution patterns as indicators of age and manner of distribution. The
distribution of languages in the great

Ubangi bend

171

ART HISTORY IN AFRICA

the eighth century A.D., and perhaps earher, communications existed with the

Middle Niger and

later (A.D.

1300+)

to

Ghana

itself.

The

object

fits

in well

with evidence that all oil lamps in western Africa derive from North African
models. The Ghanaian examples in copper alloy, cast in lost wax, were found
in this century but may be centuries old, or their prototypes may be.
Moreover, both Ghanaian lamps are of very different type (one hanging, one
on a stand), and both can be matched by Coptic examples. In addition, in the
Ghanaian copy of one lamp, the functional hinges and support of the original
type of lamp were misunderstood and miscopied, a clear sign that the Tarkwa
lamp is a copy and that the direction of diffusion was from Egypt to Ghana. It
also shows that the Ghanaian object was not intended to be used as a lamp. The
lost-wax technique, also found in the originals, would indicate that the
Ghanaian lamps were not made before either the fourteenth century, when
trade links to the Middle Niger developed, or at the earliest before the ninth

century or so, when the first archaeological evidence for lost wax in West
Africa appears - in Nigeria. We also do not know for which use the copies were
intended. The alternative hypothesis is European introduction of both types of
Coptic lamp into Ghana. They were then copied in the twentieth or in the late
nineteenth century. But this is much less likely than the hypothesis of
borrowing. Because of the remaining gaps in the evidence, the Tarkwa finds
should not be given undue weight, however. One cannot, for instance, use it as
'proof for direct Egyptian-Ghanaian links before Islam, but an historical
connection cannot be denied.
Suspicious distributions should first of all be checked for authenticity.
The Ghanaian lamps were brought from a certain Kwabena Bonda in Tarkwa
on 7 April 1936. He said they came from old graves at Attabubu in Brong
country, precisely the area where influence from the Upper Middle Niger was
strongest (Arkell 1950). It is a pity, of course, that the site of discovery was not
checked out, since dating might have been possible if the objects had been

found

in situ in the grave.

Not

so likely is the provenance of two late Isis statuettes found in Shaba.

Undoubtedly of Pharaonic times, even though late, their being found on the
Lualaba Lakes is surprising. A check reveals that their finder stopped off in
Cairo on his way to Belgium. The hypothesis that he acquired them in Cairo
and mixed them with objects from Shaba is far more plausible than the notion
that he found them in Shaba, saw nothing unusual in them, and added them to
potsherds he was going to bring to Belgium (P. de Maret: personal
communication). Such cases occur more frequently than is often realized. At
least one other case where data from Kasai and Rwanda-Burundi were mixed
has also been discovered (de Maret 1980b).
As we shall see in the next chapter, complex distributions are a feature of
all major hypotheses in art. The more complex or the larger they become, the
more they tend in fact to rely on fewer criteria. The data sometimes simply
cannot carry the weight of evidence that is deduced from them. Thus, one
scholar attempted to bring all concave, heart-shaped faces in western and
central African wood sculpture into a single distribution, surrounded by other
forms (Lavachery 1954). First the distribution is not as continuous as he claims
it to be; it is dangerous to claim that the heart-shaped faces must be younger
than the other forms, because they are surrounded by them, because the area of

172

THE CREATIVE PROCESS: FOREIGN INPUTS

too vast and too irregular. The fatal flaw, however, is that the
convex or
feature chosen is too simple. One can, after all, only carve concave,
yields a
forehead,
and
nose
a
sparing
while
carving,
concave
And
plane!
on the
heart-shaped face. The effect has therefore probably been invented many
times over. The whole hypothesis must be rejected out of hand.

distribution

is

173

CHAPTER TEN

WIDER PERSPECTIVES
FORMAL FRAMEWORKS
Having examined the properties of art works, including their creation, the next
task is to discuss the formation of complex hypotheses by which one moves
from an understanding of the history of individual or small groups of works to a
valid history of art for large regions. The core of such hypotheses is to be an
account of the succession of styles, for art history is above all a history of
shapes. Many hold that shapes have their own dynamic and point to impressive
long-lasting drifts of whole styles to prove it. Once the correct stylistic
sequence is found, technological and iconographical questions can be woven
within it, just as changing social circumstances of use and meaning should be.
Most scholars in the field have attempted to develop coherent frameworks
based on similarities of form, either over space or over time, extending the
techniques of styhstic seriation discussed in Chapter 5. In this chapter I shall
first examine such attempts and propose an alternate model of presenting
formal change over time, then I shall argue that stylistic sequences by
themselves do not lead to a history of art. It is not merely a question of placing
such sequences in the context of technology, iconography or general history.
Rather, a general historical sequence must be the fundamental framework, the
point of departure for any history of art, a history in which formal development
is only one of the elements, however essential, along with technological and
iconographical development.

FRAMEWORK BY ART AREA

A dated object in sub-Saharan art is as rare as a crock of gold at the end of the
rainbow, and chronological arrays have therefore been impossible to set up.
The situation is so pronounced that all handbooks of 'African art' had to be
content with the description of typical icons and stylistic characteristics in one
area after another, implicitly all at the same temporal level. Conventionally,
they begin with the West African Sahel and end with eastern and southern
Africa, perhaps, I suspect, because we read from left to right, from top to
bottom! The classification obtained is entirely artificial, however useful it may
be for the tentative identification of objects from unknown provenance, a
major concern to museums, dealers, collectors and scholars. Because styles
and types of objects differ greatly from one regional tradition to the next, the
system works well enough, but no genuine art historical hypotheses were
developed at all. Ordering in time was not considered; ordering in space took
its

place.

174

WIDER PERSPECTIVES
It may be unavoidable but it breeds disaster. In the
for
the erection of larger frameworks of generalization
allow
end
because complex historical hypotheses are lacking. More immediately, its
major drawback is the ethnic and spatial implications it carries as hidden
baggage. I have already dwelled on the vague and often false assumptions of

The system works.

it

fails to

equating style and ethnic group {contra Fagg 1965), the problem raised by
replications found in several ethnic groups but deriving from a single prime
work, the disparity of styles within a given ethnic group, especially with regard
to different classes of objects, for example, statues and masks, and on the
fluidity of ethnic concepts over time. Here I focus on the equation of ethnicity
to the exclusion of time.
Students of sculpture imagine the map of West and Central Africa to be
the juxtaposition of ethnic territories. Some areas are large, such as that of the
Bambara or the Senufo in the Sahel. Some are tiny, such as those of the
'grasslands' of the Cameroons, where every single town is a separate unit.
Typically, art historical maps either provide separate enlargements for the
grasslands or just list all the names. Some areas correspond to states such as the
Asante empire or the Kuba kingdom, some refer to a wider meaning of
ethnicity as when Akan appears on the map and Asante is left out. In some
cases purely geographical criteria become ethnic labels. The 'grasslands of
Cameroon' are an obvious instance. The use of Unguistic criteria makes sense
only when the languages grouped together are dialects, that is, mutually

and space,

understandable, for in that case the situation attests to frequent social

intercourse. Beyond this, linguistic criteria are unconvincing because they
refer to a common cultural heritage too far back in the past to be of use.
The spatial units compared vary, but more insidious than this is the
conviction that ethnic territory is homogeneous. All the workshops of a given
ethnic unit must make the same sort of icons, the same replications from the

prime works accepted in that unit. The corollary is that different ethnic groups
must have different styles and different prime works. None of these
assumptions needs to be true. Usually they are not. Lastly, by some sleight of
hand, it is assumed that all visual arts of an ethnic group partake of the same
features. This is not true, as a comparison between a Kuba mask and a
royal statue (Plate 8.4) shows, and this is only in the realm of sculpture.
Style is closely tied to genre and genre in turn to the institution that
commissions and uses the works.
Thus we have maps of stylistic areas. The style of each ethnic group is
characterized by the style of a 'typical' icon only. In West and Central Africa
this was almost always a statue or a mask carved in wood, 'typical' because
well known from collections and illustrations. The earliest objects stemming
from 'Tribe X' became the gauge by which all later objects were evaluated, for
these earliest objects became the visual concepts associated by scholars with
'Tribe X'. Objects acquired later were dubbed atypical if they did not conform
to the gauge, even though perhaps acquired half a century after the first gauge
had been set, the atypical works greatly outnumbered the typical ones! Such

works were then held to have been made all over the associated ethnic
area at all precolonial times and to represent a tribal style valid for a// objects in
all media made in that area, as a shorthand expressing the genius of the ethnic
group inhabiting that area. Possibilities for confusion and error are staggering.
'typical'

175

ART HISTORY

IN AFRICA

Even the most rigorous art historians did not altogether escape the
consequences. Professor F. Olbrechts (1959:29-31) is rightly remembered for
his objective approach towards the determination of stylistic criteria. In his
attempt to establish the stylistic areas of Zaire during the 1930s, he grouped the
sculpture from known ethnic provenance together, using both styhstic criteria
such as proportions and treatment of detail and the simultaneous occurrence of
from different prime works in the ethnic area. His greatest success
was the discovery of the Bull long head style, which must stem from one
workshop, or even from one hand (see Plate 5.4). One of his failures in the
same general body of data was to see that he had a large subset of typical icons
and a simultaneous occurrence of these which set that body apart from the
general mass of works included under the ethnic name 'Luba'. It was not until
the 1970s that F. Neyt (1977) disentangled this and using Olbrechts' own
methods set up a Hemba style. Neyt was himself still entrapped by ethnicity.
Typically enough, he starts his account by claiming that the Hemba were not
Luba, nor Luba-Hemba, but a separate ethnic group. That is irrelevant! What
matters is that in a series of geographically contiguous workshops, art work
was produced that differed stylistically from all other artwork. A set of
interrelated styles in the sculpture of ancestor figures was shown to exist. The
same workshops also produced some other works, such as a distinct type of
mask, not found elsewhere. But it is not at all certain that in all sculpture or in
architecture these places differed in fact from others surrounding them. The
'area' holds only for ancestor figures and to a certain extent two or three other
traditions of replication for wood sculpture. These figures may well be 'typical'
replications

for the area, but they cannot stand as a shorthand for

all

sculpture or

all

visual

art there.

Areas are best thought of in relation to one type of object at a time. Failure
this erases the spatial imprint left by the dynamic evolution of art. To
present a Kuba area, including the Kuba kingdom, or - as is now usual - add to
it some neighbouring territories and then to describe Kuba 'visual art' is an
unwarranted generalization. We saw that the distribution of double-headed

to

do

cups differs from that of cephalomorph pipes (they barely overlap!) and the

masks made out of plaiting, divination

have learned, a generation
Anthropologists
instruments and so on, all vary.
Leynseele 1978). Art
Van
and
(Kuper
areas
culture
about
forget
ago, to
historians should also reahze that such mapped areas are reifications, spurious

distribution of ceramics, types of

concrete-looking realia, which evaporate under scrutiny.

Alternatively, areas of art can be thought of in relation to a reference
group that has historical validity. Thus, the Asante capital, Cairo, Fes, etc.,
were central locations that set tastes and exercised a wide influence around
them. The total assemblage of art objects in such capitals would be the corpus
of reference and the art area then represents the area over which works were
influenced by elements of this corpus. Such art areas need not be exclusively
thought of in political terms. In medieval European architecture, there was a
Scheldt, a Meuse and a Rhine style, that is, styles by river valleys which served

and for the transmission of art. Art areas in this sense have
most part not been established yet south of the Sahara. In many cases,
the poles of artistic attraction were not dominant enough to create obvious areas
around them. Art areas are problems to be studied realistically in relation to

as arteries for trade

for the

176

WIDER PERSPECTIVES
societies over time,

and not shorthand means

to provide a classification of

objects.
to claim that stylistic areas of the classificatory type are
novices
in the field. But even this may well be illusory. The
useful to orientate
massive reductions of reality are too great. Thus, students introduced to a
Dogon (Mali) art area may never realize that much of the architecture should
not be seen in Dogon confines only, but as a variant of a body of mud
architecture typical for the whole upper Middle Niger. They can be baffled by
some statues which look as Senufo as Dogon or other works that can be
It is

tempting

confused with products from the southeast of Dogon country. These can be
distinguished but sometimes the distinction is finer than other distinctions
even within the Dogon or within the Senufo styles. Any advantage of the
system as an introductory guide is soon lost.
Classifying by art area has led to the inability to produce any wider
framework for the arts of sub-Saharan Africa, even though the need for such a

framework has been felt almost from the onset of svstematic inquiry.
Various attempts were made to enlarge areas, to group several or many of them
in larger units, but the reductionism invoked in this led to failure. Some
authors speculated on the basic distinction between pole-like sculpture and
rounded styles. The first respected the geometrical form dictated by the tree
trunk, the other broke away from it. Gradually the round style came to be seen
as naturalistic and the pole style abstract. Some sculpture that was not pole-like
at all came to be lumped with the latter because it was perceived as abstract.
In 1960, Leuzinger's book even charted the two types. Lavachery (1954)
attempted to replace the pole/rounded opposition by one featuring concave or
convex faces, and Baumann (1929), a culture historian, attempted to Unk the
'round' styles to political centralization, but none of these attempts succeeded.
There were too many cases such as the case of Ife where both pole-like (or
schematic) and round (or naturalistic) works coexisted (Willett 1967). None of
the distribution maps showed convincing distribution areas. Textbookwriters
larger

were not convinced, and continued to subsume ethnic style areas into practical
units such as 'coastal peoples of Guinea', 'Gabon' or 'East Africa' (Bascom
1973; Vogel 1981).

TREE MODELS: NIGERIA FROM NOK TO YORUBA

The

hazard any art historical

But such hypotheses could be attempted and frameworks

spatial criterion writ large implied the refusal to

hypothesis.

constructed larger than the consideration of a single style sequence. One

paradigm imagines an art form as developing from a single point of departure
whereupon drift and mutation yielded various branches that further developed
in their own right. The 'tree model' is often used in historical linguistics. A
model of this type has been elaborated for some Nigerian styles. According to
this model, the ceramic art of Nok (before 500 B.C.-A.D. 200) was the common
trunk. After transitional terra cotta sculpture found at Yelwa (A.D. 200-700),

body of sculpture, both ceramics and copper alloys, is a

body of art from the city of Ife (c. A.D. 100-c. 1500). Ife, the model claims,
grew out of the experience of Nok. Later works of Ife coming closer to modem

the next chronological

177

ART HISTORY IN AFRICA

Yoruba sculpture are thought to date from between 1500 and 1700. By then
Yoruba sculpture was established, that is, sculpture in wood, copper alloys,
ivory and fired clay practised in the cities of western Nigeria inhabited by
Yoruba speakers. From c. 1600 onwards, a large state, the Oyo Empire, ruled
much of this area. A body of art from the city of Owo was seen as an offshoot of
the art of Ife. It probably flourished in the fifteenth and sixteenth centuries.
Owo ceramic figures also show relationships to another body of art: the copper
Benin (Plate 5.7 and Fig. 5.5). We saw in Chapter 5 how a
sequence for the brass heads and the plaques of Benin was established
by which the earliest works were those that seemed most similar in technique
and in style to those of Ife (pp. 95-7). Because Benin traditions claim that the
technique was imported from Ife, the whole sequence was then seen as derived
from Ife, from c. A.D. 1400 onwards. The art continued to flourish at the court
there until our own day. Thus, from the single trunk of Ife art branches arose
that became Yoruba and Benin art. Within Yoruba art many lines can be
distinguished, Owo being one of the most prominent. Over the years stylistic
similarities were found and used as evidence to back up the proposed
sequence, and further evidence is still being marshalled to strengthen the

alloy heads of
stylistic

latest work by Eyo and Willett (1980) presents it clearly.

Figure 10.1 graphically represents the hypothesis as it now stands.

framework. The

WESTERN NIGERIAN STYLE RELATIONSHIPS*

Over the years circumstantial evidence such as the discovery of Owo art
seemed to strengthen this hypothesis more and more. But archaeology also
recovered another body of sculptures in the general area, at Igbo Ukwu, that is
totally different from any of the arts described, moreover, that is dated to c.
A.D. 800-1000 (Plate 4.4; Shaw 1970b). There is no tie between Igbo Ukwu
and Nok sculpture either. So it is now evident that the tree model is inadequate
to explain the evolution of all sculpture in ceramics and copper alloys for
Nigeria. Once this is acknowledged, other troublesome facts must be brought
up. First, there exists a body of so-called Lower Niger bronzes, which do not
belong to either the Benin or Ife traditions (see Plate 6.2). Some have
been attributed to Owo on the grounds mainly of similarities, but
others belong to one or more otherwise unknown traditions. Two figures of a
group found near Jebba and Tada on the middle Niger have been dated to c.
1300 (linked to Ife) and to c. 1400-1500 (linked to Owo?). Then there is a large
corpus of stone figures found at Esie and undated, but certainly old (Stevens

tentatively

These have tentatively been shown to exhibit some characteristics

derived from Nok but are definitely hard to place. One scholar links them to
Yoruba art, while another denies any direct link with either Yoruba or Ife.
One of the weak points of the hypothesis has been the chronological gap
between the main body of Nok and the art of Ife. Only the two ceramic figures
of Yelwa bridge it, but they are geographically eccentric and stylistically far
removed from the main Nok styles. Moreover, they are removed in a direction
away from Ife art. Then there is the fact that despite similarities, the main style
1978).

of Ife
*

See

178

is

naturalistic

Shaw

and the main

styles of

Nok

are not.

1978; Lawal 1977; Fraser 1980, 1981a and b; Tunis 1981.

To compound

the

WIDER PERSPECTIVES

-1900

-1500-1700

1400

1100

(YELWA)

NOK

Fig. 10.1

tree

model. Western

Nigenan an

700

700BC-200AD

styles

179

ART HISTORY

IN AFRICA

NOK CULTURE

\
\

Abuia

Bwari

Taruga X

nitsha

_^

Igbo

If

Ukwu

Key
I

Nok

X Nok

culture area

40

The Benin

Fig. 10.2

80

Km

site

Kingdorri core and extension

Major towns and

archaeological

sites in

western Nigeria

problem, it should be realized that Nok sculptures have not so far been
properly published and that over the vast area and long periods covered by that
term, several styles may in fact be lumped together. It is not at all certain that
the main inspiration for the Ife corpus derives from Nok. Similarly, although
Ife objects have been found in Benin and vice versa, Benin brasses probably do
not derive mainly from Ife. The dates can be construed so that work at Benin
partially overlaps with that of Ife, but the stylistic filiation is unconvincing
(Fraser 1980, 1981a; Tunis 1981).

Owo

and late Ife are being

and ceramics. Part of the
difficulty here stems from the use of different media, namely wood and stone
sculpture. A major problem moreover is to decide whether perhaps the Esie
stone sculptures may not be closer to later Yoruba wood sculpture than the
latter would be to ceramic work from late Ife or Owo. Whatever the general
proposed links, the tree model certainly fails in the attribution of only a single

The

links

between Yoruba work and both

better established especially for

180

work

in metal

WIDER PERSPECTIVES

Plate lO.l

Pan of a figure. Poiien\Jemoa{\igena). LagosMuseum.

194^

at the

Tsaum

many

others

found

mine. Dated between 500 B.C.

and

.\.D.

Phoio Bniish Museum. Height I4im.

200. The object

is

usually labelled

northern \'igena, usually during tin mining. 'Nok' objects share a

eyeballs, but oiherivise belong to

many

few

'Nok

traits

Foundm

culture', as are

such as the pierced

different styles

Yoruba art. Owo, Esie and Ife should be considered as some of its
Moreover, every sixteenth-century Yoruba city was also at the root of

'ancestor' to
roots.

later art.

Beyond this, the model contains three major flaws. First, it tends to
generalize from a set of specific works, the ceramic and copper alloy figures, to
all of sculpture, which is unwarranted. Benin sculpture in wood or ivory differs
markedly from the canons of Benin copper alloy heads, for instance, and
Yoruba characteristics of statues of the earth mother, used in the Ogboni cult,
cannot be extrapolated to stand for all Yoruba art.
181

ART HISTORY

IN AFRICA

Plate 10.2 Crozuned head. Zinc brass?

12th

and

early 16th

c.

leading chiefs also, as

182

Ife, Nigeria. British

Remains undated. Such heads

we know

Museum. Height 21 cm. Acquired 1939. Made between

the

are interpreted as royal portraits, but they could be portraits of

nothing about Ife's government at that time

WIDER PERSPECTIVES

Fig. 10.3 Head. Ceramic. Use unknown, perhaps in shrine.

Dated by

distinguishes

it

from similar sculpture at

Secondly,

which are

it is

Ife.

Owo,

Nigeria. National Museum, Lagos. Height 17 4cm.

The roundness of volume and

After T. Shaw 1978:185

association (date of the excavation), 15th

c.

the liveliness of expression (lips!)

evident that the tree model presupposes features of change

valid in linguistics, but not for art.

The model

refers to a single

totally integrated system (the language code), and attempts to establish a point
in time, the node, at which languages split. Since the visual arts are not a totally
integrated system, no such node can be found. Indeed, prime works are
created out of multiple influences, whereas the tree model presupposes a single
ancestor. The model takes into account only genetic links, leaving aside all
later mutual influences. Again because such influences are of the greatest
importance in the shaping of prime works in art, the model cannot be used to

express changes in design and composition.

The last flaw is the belief that because one phenomenon follows another
one, it must derive from it. In fact, Nok, Ife, Benin are a succession of styles
over time and nothing more. The resemblances adduced between the styles of
the representational ceramic and copper alloy sculpture are not so
overwhelming as to prove filiation. On the contrary, intensive search for them
merely establishes some influences for a few discrete features. It would be
if it were not so, given the locahzation of Owo, Ife, Benin and
Yorubaland and the widespread occurrence of Nok ceramics almost a
millennium before the appearance of the other styles. One can find some

surprising

183

ART HISTORY

IN AFRICA

between Nok work, later ceramics around Lake Chad (So) and
even between Nok and Igbo Ukwu. Just the passage of such a vast amount of
time, coupled with the huge territorial expansion of Nok styles, would lead to
the expectation of such similarities.
To sum up: multiple centres for the production of ceramic and metal
sculpture are old in western Nigeria. An urban or semi-urban way of life had
developed there since the onset of this millennium and every city became a
centre for the production of sculptures in various media. Multiple
relationships and overlapping influences should be taken into account, hybrid
inspiration being very common in art. The term Yoruba should best be
dropped, and the hypothesis restated in terms of types of art and of cities seen
as clusters of workshops. It then becomes evident that the term Nok refers to a
similarities

Nok never was a city. It is a label applied to a set of traditions

seven hundred years and more in duration over a huge area of Nigeria mostly
between the Benue and the Kaduna rivers, north of the Niger. Truly
satisfactory links with later work, as at Ife, could be found only after the

different reality.

Nok pieces have been published by locale and the relationship

between the localities has been worked out. Thus the vague term Nok should
be replaced by specific place names such as Kafanchan, Jemaa, Nok, Bwari,
Abuja, Taruga, Nasarawa, Udegi, Katsina Ala, Yola, etc.

clusters of

ultimate failure, the research hypothesis in the shape of a tree

model had a great virtue. It attempted to take time into account, whereas none
of the previous research had done so. But how then are we to visualize the
overall development? The evidence allows us to state the following: from 700

Despite

its

500 B.C. ceramic sculpture was widespread in the Nigerian middle belt. It
was produced in many centres and represents several stylistic sequences, as yet
to the west
still unknown, but all labelled Nok. Ceramic traditions continued
of this area at Yelwa and to the north around Lake Chad where a still insecurely
dated ceramic art flourished from the beginning of our era onwards. It is given
the label So and represents the production of several centres as well. A
progression of styles lasted over many centuries, perhaps only ending by A.D.
to

1600.

southern Nigeria, the later Yorubaland. Ife, one of

those, perhaps the oldest, produced ceramics, metalwork and stone sculpture
between 1100 and the seventeenth century, although the dates are still
provisional. As other towns developed in Yorubaland or to the southeast, they
such
also developed such arts. Indeed, east of the lower Niger, evidence for
been
have
must
but
town,
a
been
never
arts is found at Igbo Ukwu, which had

Towns appeared

in

an important trading emporium. Igbo Ukwu clearly did not influence

the art of Ife at all. Yoruba cities, however, probably all influenced each other,
Ife was
to unequal degrees as the prestige or power of each waxed and waned.
of Ife
those
from
differ
traditions
ceramic
Its
one of them, Owo was another.
can
source
third
The
source.
third
similar
or
a
Ife
by
influenced
clearly
are
but
be eliminated if the accounts that Ife was the wellspring of Yoruba cities and
monarchy are accepted, as most historians do, although it may well be that
at least

such traditions refer only to relatively late dates before A.D. 1600 and not, say,
to the situation around 1300 or so.
Benin slowly grew among other settlements in the forest, became the
leading town by 1200, and developed its arts there. It later established
184

WIDER PERSPECTIVES
connections with Ife as finds in both Ife and Benin show. Benin had
connections with Owo based on evidence of similar iconographic motifs, but
this type of evidence is weaker, for it does not show Owo objects in Benin city
and vice versa. Still, the proximity and the later subservience of Owo to Benin
allows mutual influences as a reasonable hypothesis. The sculptures found at
Jebba and Tada as well as others for which a locality of origin has not been
documented, show in fact how far such influences from Ife, Owo, Benin, could

then be found.

account does not consider any media other than clay or metal.
Yoruba sculpture used stone, wood and ivory. The undated
stone sculptures from Esie may have come from a rural background, since they
are not associated with any city. The technique of carving is similar to

So far,
But Esie or

this

later

woodcarving techniques and we may suspect that woodcarving had been

practised since remote times in Yorubaland because the Ifa board (Plate 1.1)
made in the 1650s, or earlier, shows a refmed technique and a well-developed
genre, comparable to later Yoruba work. Note in passing the irony of using the
term 'Yoruba' for these arts when the object was found at Ardra, a
non-Yoruba-speaking area! Some of the ivory carvings in Yoruba styles and
the carvings made in Benin for Portuguese in the sixteenth century point to the
same inference. If ivory was carved, why not wood? Yoruba sculpture must be
the continuation of the earlier sculpture that was thriving in the older Yoruba
cities, places such as Ife, Owo but also Old Oyo, Ijebu Ode, Ilesha and perhaps
others, all thriving in the sixteenth and seventeenth centuries. The whole
complex of interrelated workshops in the earlier cities and rural traditions such
as Esie or others is ancestral to that of later Yoruba art. A link between a work
from Tada, attributed to 'late Ife' and a well-known political cult, the Ogboni
of later cities, consists in an iconographic detail: the Ogboni salute. But there is
no clear hneai link between later Ogboni brasses and Ife styles. That indicates
how difficult it will be to unravel precise strands of development. Moreover,
skeuomorph development also occurred because some features on ceramic and
metal art from Ife are found on wood in modern Yorubaland. Still, the whole
tradition of all interrelated Yoruba styles, and there are many, must be linked
with the whole tradition of all known and unknown sites in the area in the
sixteenth and seventeenth centuries. That means that the recent sculpture is a
descendant, not only of Ife, nor even of Ife and Owo, nor even of all early
Yoruba cities, not even of Yoruba cities and rural workshops, but also, in part,
of the Benin world of sculpture, if only because in those centuries Benin was
the overlord over southern Yorubaland and hence the point of reference for
good taste. The catalogue entry for the Ifa board from Ardra mentions the
supremacy of Benin. Benin art must also be counted among 'Yoruba' ancestry
in sculpture because the city was then interacting with several Yoruba cities
and it is the whole of the interacting network that is ancestral to later Yoruba
art. Given these considerations, it becomes obvious that tree models will not
do. A graphic sketch of a more appropriate model is set out in Fig. 10.4.

AN OPERATIONAL MODEL: STREAMS OF TRADITION

Any

model for the evolution of form must take into account multiple
borrowing and continued mutual influences between neighbouring

valid

origin,

185

ART HISTORY IN AFRICA

I.

Time and place where

Nok

sites (many)

art

appears.

WIDER PERSPECTIVES
while allowing also for renewed influence of art works crafted
generations ago in the same area and perhaps in the same tradition, as in the
case of Benin work from the eighteenth century imitating objects from the
sixteenth (Ben Amos 1980a:30 34-7). At the same time, the real continuity in
the art of any 'style' should be recognized. For small scale drift is also a major
component of change. The evolution of style is like a large river, receiving
water from neighbouring currents and giving off water to other currents, but
nevertheless pushing a large mass of fluid downstream in time. Rivers can take
on tributaries, lose tributaries, meander and fork, rather like the multiple
Styles,

outlets in the delta of the river Niger fed

falling into the delta, or they

by the main river and by smaller rivers

can flow smooth and straight in the same channel.

and yet there are ancestors and there are

of streaming traditions.
model
continuing streams. We
With such concepts it is evident that there are not enough data on art,
especially south of the Sahara to detail all the multiple interactions that were
taking place. We must always assume then that we have more lacunae than
data, whereas the tree model implies the contrary. When chronology indicates
a broad synchronism of streams, mutual influences of such traditions on each
other should be expected and found by stylistic comparison without prejudice
to other links that also can be substantiated by style.
As an illustration we can survey the situation of sculpture west of Nigeria
from c. 1100 to c. 1600. Three centres can be recognized then in the upper
Middle Niger area: the environs of Jenne and of Bamako produced terra cotta
in human form (De Grunne 1980; Mcintosh and Mcintosh 1981;Bedaux,gra/.
1980), while the environs of Goundam yield varied work in metal and ceramics
(Davies 1967:262, 265). By 1 100, wood sculpture was practised, and some has
survived near the cliffs of Bandiagara, not far from Mopti and the Jenne
concentration (Bedaux et al. 1980; T. Northern, personal communication).
Stone sculpture, undated so far except by stylistic association with
Afro-Portuguese ivory (Dittmer 1967; Paulme 1981; Person 1961; Atherton
and Kalous 1970) appeared further south near the upper Niger valley in Kissi
country and then towards the coast of Sierra Leone. From the latter area,
ivories were exported to Europe fromc. 1500 toe. 1600. Contemporary written
sources tell us that stone and wood sculpture were practised in the Niger bend
during the fifteenth century and wood sculpture at the capital of the MaU
Empire a century before, and probably since at least 1 100, but no works from
these areas have yet come to light. Finally, to the southwest, ceramic and metal
works from modern Ghana, south of the big bend of the Volta, are known to
have been created from 1400 onwards although pottery heads and figures are
extant only from c. 1600 and later (De Grunne 1980:137-9).
Using the stream of tradition model, we must presuppose that each style

There are no

single ancestors

need

centre influenced its neighbours. The multiple ateliers that can be recognized
in the Jenne area are clearly all interrelated and form one stream. Despite
obvious differences, there are also clear links with the Bamako area, from
where we have mainly a single rivulet - perhaps because so few works have
been recovered there. Links with Goundam are much less evident, except

perhaps for metal work. A few Jenne pieces can be put side by side with work
from Bandiagara cliffs (Tellem) and one or two from Bamako strongly recall
later

wood carving tradition to the southeast:

the traditions of Upper Volta and

187

ART HISTORY

IN AFRICA

Plate 10.3 Bust. Ceramic. Inner Niger delta, exact

mound unknown. Mali.

Undated. Similar figures have been dated between the Ilth and 16th

188

c.

Private collection. Height perhaps 20cm.

WIDER PERSPECTIVES

Fig. 10.5 Disinhution of four-legged pots along the Upper Niger before A.D. 1500

Ivory Coast labelled 'Senufo'. That there really were contacts is proven by the
presence everywhere along the Niger and at the Bandiagara cliffs of a
well-defined ceramic vessel with three or four legs, found as far downstream as
Gao and as far upstream as the region around the capital of the Mali empire,
Niani (Fig. 10.5). Links between Jenne and the town of Begho in the lands

beyond the lower Volta have been proven from 1350 or earlier onwards by
ceramic evidence and by the system of gold weights. There is, as yet, no direct
proof of links between this cluster and the cluster of stone and ivory work from
the upper Niger and Sierra Leone, itself divided into several interrelated
streams, but no detailed research has been undertaken and the differences in
the media of the art works are obstacles barring easy identification of mutual
influences.

1600 is compared with the situation earlier, some

Thus, metalwork from Goundam, which had
correspondences with metalwork from Mopti in the Jenne centre, is
comparable with jewellery later worn by the inhabitants of the Sahara, while
other pieces, mostly bracelets from Goundam, are directly comparable to
bracelets on the lower Ivory Coast and in Liberia from the nineteenth century.
This should not be surprising as lost-wax techniques for dealing with copper
alloys diffused from northern Africa, presumably over the upper Middle Niger
Valley (a casting mould near Mopti dates from A.D. 1000-1200, Bedaux et al.

When work

continuities

after

appear.

189

ART HISTORY

IN AFRICA

Plate 10.4 Masked persons. Watercolour by Frobemus. Mosst, Upper Volla. Frobemus Institut. Original 1907-1909,
entitled:

in masks' Probably chiefs of the land, representing the authority of the aborigines before the Mossi
Birdmasks are ancient in West Africa being attested since the 1350s

'Mosst priests

conquests.

1980: 146) to the south from the coast of Ghana to Liberia. As far as sculpture in
is concerned, the later styles of the Bandiagara cliffs belong to a stream
that is in part the direct continuation from the earlier Tellem works (see Plate

wood

some outside influence and local innovation. In Sierra Leone and in

the Kissi area, the most significant feature however is a pronounced break
between the ivory and stone work on the one hand and later carving in wood on
the other. Very few influences cross that divide. But later wood carvings,
2.5) with

especially

from southern Sierra Leone, Liberia, the uppermost Niger, and

portions of the Ivory Coast, share at least one icon, the bird-shaped mask
(Bravmann 1974:47) that was reported for Mali by Ibn Battuta in 1352/3; even
the red colour of the beak has been preserved in some cases (Plate 10.4). There
must therefore be some continuity between these styles and wood carving in
the 1300s, probably in southern Guinea. The obvious similarity between
Afro-Portuguese ivories and styles of carving from eastern Liberia and the

Ivory Coast - namely, that carving was done in the round, with a miniaturist's

190

WIDER PERSPECTIVES
not a case where ivory work
think that the later styles flow
from a stream of carving in wood which also influenced the ivory work from the
sixteenth century. And finally we begin to perceive that different clusters of
ateliers in the Sahel made products c. 1900 that still share some features with
attention to detail and great care for the finish,

gave

the

rise to

wood

Bamako

sculpture. Rather,

is

we should

centre ceramics.

We

have in this simplified sketch looked for both synchroneous

relationships and filiation. We find clusters linked over time and for any given
time amongst themselves and we conspicuously leave the lacunae as the main
theme! Nevertheless, this does yield an operational framework for research in
which all the sculpture of West Africa, west of the Volta and the Niger bend
can be placed. Whilst it is evident that there are no direct relationships at all
between the traditions we dealt with and either Nok or any of its cluster, future
research probably will find transitions between the two, traces of which can be
expected along the coast itself or from northern Ghana eastwards to northern

Dahomey.

We

have intentionally limited ourselves to sculpture.

If

we

include

most relevant feature to stress is the existence of an

advanced mud architecture in the middle Niger from the last centuries B.C.
onwards. The area of this architecture stretches from there over the Upper
Volta, northern Ghana, Togo, Dahomey (Benin), to Nigeria. But the study of
architecture is still in its infancy and we can say very little about it, except for
the obvious: the mud architecture of the middle Niger belongs in that
architecture, the single

environment!
At this point the informed reader complains: you have illustrated the
approach by using the one area for which somewhat adequate archaeological
information exists. It at least gives you some hold over time, some straw to
make bricks with. No doubt you could make a similar demonstration for stone
architecture in southern Africa from c. 11 00 onwards! Indeed, we can and the
Zimbabwe ruins are just one stream in that delta of data. What if archaeological
data are absent, however, or so sparse as not to make a difference? Is that not
the situation in most of central and eastern Africa?

Such

a perspicacious reader

history in Africa as elsewhere.

is

right.

Archaeology

is

as crucial to art

Who would have foreseen a ceramic sculpture in

Natal and Transvaal from c. A.D. 500 to 800, if the trowel had not uncovered it
(Maggs and Davison 1981; Plate 10.5)? But archaeology without an
operational framework of interpretation remains barren. The stream of
traditions model provides that. It allows us to beware of quick conclusions, for
instance that the wood sculpture in the form of heads topping poles and the
stylistic features of the face in central Angola and along the Kwango and
Middle Kasai must directly descend from the style of the first millennium pole
found at Tumbica on the upper Kwango. That is tree-model reasoning and the
being correct are abysmally small.
us however to think in terms of real time
depth, to see genuine links between space and time. At present we cannot
reach much beyond the nineteenth century for most of the sculpture of
southern Zaire, but we can perceive affinities over large areas and think in
terms of clusters of streams of tradition. Thus, for instance, icons as caryatids
supporting stools are a typical theme from the Kwango to Lake Tanganyika.

chances of

it

The same approach encourages

191

ART HISTORY

IN AFRICA

Plate 10.5 Head. Ceramic. Use unknown. Lydenburg, Transvaal. Natal Museum. Height 38cm. c. .4.D. 56*0. Since
photo was taken the mask has been restored with an animal crest on top of the head. Compare cover African Arts,

this

1981, 14

192

(2).

One

of seven heads and the largest

WIDER PERSPECTIVES
They do not prove that central Angolan and eastern savanna traditions share
the same ancestry. But they do point to undeniable mutual influences. The
study of spatial distribution tells us unmistakenly that the history of sculpture
here is quite complex and reflects many different historical processes of drift
and mutation.

THE FAILURE OF FORMAL FRAMEWORKS

By themselves formal sequences cannot provide

a satisfactory

framework

for

sequences are dated. A

general social and cultural history is crucial to a satisfying understanding of
relationships in art history. Thus, when we deal with western West Africa, it is
crucial to take into account that ancient Ghana, ancient Mali and Songhay were
empires flourishing there; that Islam penetrated the area slowly, first at the
courts, later in Songhay after 1493 as a state religion; that trade for gold first,
kola later, linked the forest to the savanna to the desert and the need for salt
linked the coast to the southern Sahel and the Sahel to the deposits of the
Sahara. Our sketchy model of the evolution of sculpture becomes convincing
only when we put it in full historical perspective. We account for breaks in
tradition (Sierra Leone, middle Niger), for links over long distances (middle
Niger-Ghana south of the Volta), for continuity (the cliffs of Bandiagara), for
centres of production (Jenne to 1600). The skeleton is being covered by flesh.
The need for an overall historical framework is dramatically shown in
cases where we do not have large numbers of art works from early sites. For
eastern Zaire we know that a sophisticated technology of iron existed from c.
A.D. 800 onwards and some figurative work in fired clay was then practised (de
Maret 1977). We know that by AD. 1000 copper currencies emerged there
(Bisson 1975), and trade from hand to hand was linking the area to the Indian
Ocean. By then, or not much later, we know that chieftaincies had appeared
and that some of their emblems are those of later Luba kings, while some of the
customs, such as the way in which teeth were filed, were to survive until 1900.
The whole cluster of the arts in and around the later Luba empire can be

an

art history

because they are too limited even

progressively disentangled as

if

we come nearer to the present.

We know that we

have to allow for long time depths, and yet we can place the few early works in
one overall valid framework.
Art works should be fitted into general history with greater care than has
been the case in the past. For instance, too much is still made of migrations
carrying icons and iconic tradition with them over long distances. There is not
enough consideration of links between the sudden appearance of novel prime
works and the demands by new social institutions. There is not sufficient
awareness of leaving enough slack in interpretations for the unknown, even if
rules of evidence are scrupulously followed. But it is possible to build up a
tentative art history and it is imperative to break away from the flat map of
styles to include the temporal dimension.
The immense value of the general historical framework is most apparent
where it is lacking. Thousands of specimens of rock art in southern Africa and
in the Sahara cannot make up for it. In the end, and speculation apart, we know
very little about rock art (Lajoux 1977; Willcox 1963; Vinnicombe 1976;

Woodhouse

1977, 1978;

Rudner and Rudner

1970).

Even

if in

the future precise

193

ART HISTORY

IN AFRICA

and chronologies are worked out, most questions will still remain
begging, and only general archaeological finds of stone tools, other utensils,
perhaps the remains of domestic and wild animals, and certainly living sites
can alleviate the barrenness of a style classification in a total void. Even the
iconography of such works can really only be enriched by finds relating to the

distributions

way of life.
In contrast, where the general historical evolution is well documented, the
formal sequences of art history, as well as changes in the iconography and the
technology acquire a rich significance. In northern Africa, the conditions
general

obtain.

We have enough historical background and enough dated works of art;

at least in architecture. Stylistic similarities find their

history.

The Aghlabid

explanation through

mosque of Qairawan as an expression

and with means they could afford, as they had

rulers rebuilt the

of their legitimacy in the area

reorganized Tunisian society around their dynasty, its clients and its allies.
They then commissioned the Zaytuna mosque in Tunis, which may have been
built by the same architects and was certainly built using the same type of
materials, the same techniques, the same standard elements. This was to
impress the urban population, which was still impregnated with Christian
Roman and Byzantine culture. The differences in the domes of Qairawan and
the Zaytuna are telling. The first is a novel creation, but inspired by the domes
in Iraq, home of the Caliphs. The second borrows some features from the

Byzantine architecture still visible near Tunis (Lezine 1966). The

sequence of Fatimid portals, which we explored in Chapter 8, makes sense
only with the political and religious background behind it.
In Nubia, the historical background is not as rich but the wall paintings
provide a sturdy chronological sequence and the iconography is well
understood. The architectural remains are not as yet dated with sufficient
precision, but their sequence is clear and can be linked to the general historical
evolution. Even the ceramic sequence, which is well dated, can be tied in with

Roman

the other arts and with the social and cultural history. These arts as expressions
of culture vary with cultural changes in religious dogma, with social change in
the relative positions of monasteries, ordinary clergy, rulers or governors, and
the relationships between both the ecclesiastical and political establishment on
the one hand and the mass of farmers, fishermen, stock keepers and even
traders on the other (Adams 1977). The sources complement one another and
allow an overall historical hypothesis to be worked out.
Ethiopian painting is much less well dated, although the main sequences
are evident. Far too few works as yet are dated at all, but wall painting at least
can often be tied in directly with architecture. We also lack enough dated
buildings, however, despite the hundreds of 'medieval' monolithic churches.

But the general

historical

background, the economic, political, social and

known from the fourteenth century onwards,
This means that when more dates become available

cultural forces at work, are well

and in outline, even earlier.

for works of art, it will be easy

to place them into the general picture. Even now

successive formal sequences and successive changes in the
can measure Ethiopian
iconography of painting into more general change.
reaction to Muslim advances in the sixteenth century both in the iconography
(condemning Islam) and in stylistic evolution (dropping some Muslim motifs,

we can

securely

fit

We

but adding others).

194

We see the impact of the Catholic missions on the evolution

WIDER PERSPECTIVES
of paintings depicting the Virgin Mary in the seventeenth century, and we
begin to measure the cultural impact of the Empire's collapse a century later in
the paintings after 1750.
To conclude: techniques that yield stylistic sequences or that identify the
shape of prime works and sometimes archetypes among prime works are
necessary. But they deal with a development of form that does not lead directly
concerned with
to larger and deeper understanding, mainly because they are
evolution of
autonomous
the
documents
the least part of art history: that which

form and ignores the forces that bring about new icons, new genres, new
grouping by
'traditions'. Grouping by area and by form leads nowhere, while
area on geographical or linguistic criteria neglects formal variation and igno;:es
socio-cultural matrices. The tree model is moderately fruitful because it draws
attention to time, but

it

remains flawed because

it

considers only drift, not

multiple influences, nor radical innovation. A stream model comes closer to

for the
reality and takes unknowns into account, but it too cannot account
Only
innovation.
radical
explain
and
document
it
reasons for change, nor can
the reconstruction of the general social and cultural history can be the proper
framework in which to establish valid art history. In turn, as is argued in the

next chapter, art history

is

an essential ingredient to any history that claims to

be general.

195

CHAPTER ELEVEN

ART IN HISTORY
History without works of art remains bloodless, unreal, to me - something that
could have been, but never really was. Yet today art history has no place in
general African history. None of the general overviews, not even the
eight-volume-long series of the Cambridge University Press and the UNESCO
general history of Africa, consider art as much more than a source of
illustrations. The main journals carry very few articles on the topic, and at least
one major school of historiography is hostile on principle, as it does not
perceive a meaningful place for artistic expression in its explanation of history.

We

cannot

assume

that

art

is

unimportant

in

general

historical

reconstructions, whether in Africa or elsewhere. It is too often dismissed as an

must therefore argue the case for its relevance.
irrelevant epiphenomenon.

We

Let us begin with an example.

STONE MANSIONS AT LAMU

Lamu

is

town

in

Kenya with

IN

THE EIGHTEENTH CENTURY

a long history. Built near

one of the

earliest

trading settlements on the coast, it only reached its zenith almost a millennium
later, in the eighteenth century. At that time the coast of northern Kenya was
under the control of either Lamu or Pate, which were vying with each other for

predominance. Trade with lands across the Indian Ocean was brisk and much
wealth flowed into the towns. A number of leading citizens could afford to
build stone mansions and live in them (De Vere Allen and Wilson 1979; Spear
1981:89-96).

Stone houses had been a feature of Swahili towns since the thirteenth
century at least, but usually represented only a few buildings within a
settlement built of wattle and daub. When prosperity was great, however, the
majority of the buildings came to be of stone. So it was with Lamu in the 1700s,
where mansions from that time are still standing, albeit mostly in ruins. They
were the badge of civilization in the eyes of their dwellers and of the other
townspeople. They dramatically separated the well-to-do from the vulgum
pecus and people Uving in them also adopted different ways of life in matters of

making a living and even entertainment as well as in

and demeanour from the lesser folk. If they fell on hard times and
could no longer keep up their mansion, they left their houses and with them the
behaviour, mannerisms, tastes and predilections of their social layer. They
usually moved to another town and merged with the less fortunate (P. Romero,
personal communication). The stone house, then, was the crystallization of
religious observance,

dress, speech

status in

196

Lamu

society.

ART

IN

HISTORY

Stone houses occurred in blocks, or wards (mfaa), each block being under
the authority of the head of a widely ramified family. The architectural
evidence favours the notion that marriage was matrilocal so that women in a
ward were closely related and visited each other within the mtaa without
having to appear in the streets, something that was frowned upon in this
Muslim society, and something that set their status dramatically apart from the
lot of poorer folk.

N
5
1

Lamu. Plan of a one-storey dwelling. 18th c. a porch) b inner porch;

fouler living room; g inner living room; h women's quarters;] inner bathroom; k
After De Were Allen 1979:diag. 3

guestroom; d courtyard; e front

Fig. ll.l House,

toilet;

nyumba ya kati; / kitchen area.

Most houses were single storey, but two-storied buildings also existed and
further distinguished the wealthiest people from the others. The average single
storey house consisted of a porch with solid benches on the side and an inner
porch leading to a courtyard (Fig. 11.1). Behind the porches was the guest

room, the only room with windows, barring only ventilation holes in the toilet.
One side of the courtyard served as a kitchen area with a toilet nearby, the other
side was walled off. The bulk of the house lay on the back of the yard and its
fagade was oriented northward away from the rain. The block consisted of an
outer living room, an inner living room, the harem, an inner bathroom and the
197

ART HISTORY IN AFRICA

Kenya. 1 8th c.
Plate 11.1 Inner wall of Lamu mansion. Coral masonry, plaster. Inner walls, doorway, niches. Lamu,
plates
The niches vary in height, depth and slam for perspective. They give an illusion of depth. Brass lamps and Chinese
or bowls were set in them

kati, 'the inside room', at the back of the house. The Uving rooms
received light from the courtyard through large apertures. They were densely
furnished and served as work rooms as well as bedrooms. The harem only had a
single entrance that could be closed by a door, and this opening was carefully
out of alignment with the wide apertures from the courtyard to the living
rooms. The harem was the most decorated room in the house. The whole wall

nyumbaya

facing the inner living

198

room was covered with panels of

wall niches.

Such

ART

IN

HISTORY

panels also occurred in smaller numbers in the inner living room, on either side
of the harem doorway. The nyumbaya kati, 'middle room', at the back was the
only one, however, that was left completely undecorated with stuccowork,
which occurred to some extent everywhere else even though the decoration was
never as rich as in the harem. More complex arrangements characterized the
storied houses. There, as in the single storey buildings, the greatest care was
taken to isolate visitors from the inside of the house and even to separate clearly
the upstairs housing unit from the downstairs one. In all houses privacy was a
prime consideration, not only in order to isolate women. Thus stone houses
allowed a larger number of people including servants to live together with an
acceptable standard of privacy, as compared with those in mud and thatch

houses.

But houses were not built only with daily living requirements in mind,
they were especially concerned with marriages and funerals, and in part births.
The nyumba ya kati served to lay out the dead and take them out of the house
through a hole in the back wall. Such rituals are so alien to Islam that today's
informants vehemently deny them (P. Romero, personal communication).

The arrangement of the living rooms and the harem, with their bays and doors
and decorations, formed the proper stage for wedding ceremonies. First-class
weddings were the supreme affirmation of a family's standing and of the
groom's status and responsibilities at large. The stucco work, the niches, and
other details were all arranged to display the wealth of the families and form a
proper backdrop to the festivities, which might last a week. The whole house
or house unit itself was usually new then, a present to the newly weds, a
building planned and prepared ever since the bride was born. The high point
was the revelation of the bride on a bed in the harem against a wall space
expressly stuccoed for effect.
The study of the eighteenth-century house can be pursued in many
directions. It was the continuation of a tradition in which plans and decoration

can be traced over time. Plaster decoration, for instance, replaced earlier coral
rag carvings, much too cumbersome and expensive for the greater numbers
who in that century aspired to a stone house but who were less wealthy than the
smaller number of people belonging to the elite in earlier times. The details of
decoration, of furniture, of imports such as ceramics, rugs and hangings, show
a fme balance between imported items from all over the Indian Ocean and from
China and items made on the coast in styles that were similar to, but diverged
from, similar work in Islamic countries. These echo the results obtained by a
study of the Swahili language. Many loans, yes, but fundamentally a strong
local tradition. Houses can also be put in the context of social life. They were
perhaps the most sizeable goods forming part of an estate, they were grouped
in blocks expressing the material wealth of the block unit. They were the
product of a great expense of labour, some unfree, and their owners were
served by slaves. Details recalled links with mosques and with good Muslim
traditions as in the small arches recalling mihrabs and even in their orientation
towards Mecca.
Life in such houses could well be visualized. And they give concrete shape
to the images of poetry from the age; describing both their eclat and - for the
neighbouring town of Pate - their later ruin.

199

ART HISTORY

IN AFRICA

How many wealthy men have we not seen

Their Ughted mansions glowed with lamps of brass
.

And

crystal,

till

night seemed like very day;

Their homes were set with Chinese porcelain

And every cup and goblet was engraved
While, placed amid the glittering ornaments,
Great crystal pitchers gleamed all luminous.
The rails from which they hung the rich brocade
Were made - I swear by God, Source of all Wealth Of teak and ebony, row upon row of them.
Rank upon rank with fabrics hung displayed.
The men's halls hummed with chatter, while within
The women's quarters laughter echoed loud.

The noise of talk and merriment of slaves

Rang out, and cheerful shouts of workmen

rose.

.*
.

How can we fail to grasp the gulf of class difference in Lamu, the
implications of patrician status, the interests patricians represented in city
government, the aspirations of workmen and slaves? How can we fail to grasp
the ambiguity of the proud and haughty, however pure, devout, rank
conscious, and Muslim, but still tied to their menials by profound cultural ties?

The hidden nyumhaya

kati, the spatial

arrangements in the house for deUvery

same un-Islamic convictions also held by commoners.

and
In every generation some of the exalted fell to the mud and thatch class, just as
some of the humble rose. How can we fail to grasp the stakes when we see in the
house and its trappings the prize defended by its holders and longed for by the
birth reflected the

poor?
Swahili houses testify to the material culture, the economy, social and
life on the Lamu islands in a more direct way than other sources,
whose full meaning becomes clear only when confronted with this evidence.
The importance of these houses as primary sources is clear when one reaUzes
that historians until quite recently could not see any link between Swahili

cultural

and the civilizations of the villages around and behind the towns,
civilizations that were those of the mud and thatch crowd. By examining
material culture, J. de Vere Allen has been able to reinterpret basic
reconstructions about the past of the East coast. He fu-st showed that neither
the houses, nor their furniture, are simple copies or imports from Muslim
lands around the Indian Ocean or the Red Sea. He was thus led to stress the
civilization

originality of Swahili culture.

A study of the

stone houses as parr of cities then

and 'lower class' Swahili formed a

element obviously culturally related,

led to the insight that both 'upper class'

single society with the lower class

sometimes almost identical with the neighbouring 'tribes' inland. The history
of East African towns, which had been totally divorced from their hinterland
and cultural heritage, could then be rewritten. It remains doubtful whether his
insights could have been achieved in any other way.

Quotation from the translation of a/ Inkishafi by

200

J.

de Vere Allen, Nairobi 1977, pp. 63-4.

ART

IN

HISTORY

ART OBJECTS AS SOURCES

The

relevance of art history to history in general

level the use of art objects as sources for history

is of a double nature. At one

must be assessed. At another

the relevance of art history proper, that is, an account of the history of style,
iconography and technology, must be considered in relation to history in
general.

Art objects, as

all

objects, are traces of the past,

and hence are sources.

They are tangible and more or less j)ermanent. They are direct witnesses to the
time when they were made and used and having been of practical use they do
not carry any bias, other perhaps than that induced by the hazards of survival.
Collections of art works should be as fruitful as collections of objects recovered
from sites by archaeologists. After all, quite a number of art works are
archaeological finds whilst others are classified by methodologists as
monuments, a species of archaeological material that was never buried! They

except in some narrative compositions,

narrative works do not even relate to
and
history. Painting in the oikoumene refers rather to sacred history and among the
plaques of Benin only some purport to portray an historic event. But it is true
that great figures of the past are represented in art.
Representative art is a testimony to what it depicts. But we have seen that
problems of interpretation are often thorny and presuppose that the artwork be
accompanied by another source, oral or written or epigraphic, which explains
its rendering. Nevertheless, some evidence is direct enough to be grasped
immediately. Often historians do not quite reahze how much a given body of
art work can contribute as a source. A good illustration is given by the
goldweights of southern Ghana and neighbouring parts of the Ivory Coast.
They are first a record of weighing systems, related to North African and
European systems. They also are a mute witness to the practices of the gold
trade, the institution for which they were used. A repertoire of goldweights
encompasses virtually all utensils common in the Akan world from c. 1600

testify to situations, rarely to events,

in Africa they are rare.

onwards.

We find the various forms of agricultiu-al tools, fishing implements,

tools for the craftsman,

fiu-niture

Most of the

and

weapons

(for old shields, see

Garrard 1980:113),

utensils for cooking, representation of dress, ritual objects,

appearance of chiefs, traders, farmers, women of varying

whole little world of people engaged in
their daily rounds and on their exceptional activities. Some 300 000 figurative
weights have survived out of a total estimated at three million. One third
consists of animals and plants and thus shows us the natural world interpreted
by the Akan mind. A further one third consists of horns, knots, shields,
daggers, hammers, axes and fans. Still that leaves yet another 100 000 to
represent other artefacts. An estimated 8 000 human figure weights show man
engaged in all sorts of activities. Here some portrayals may be unique,
according to Garrard. Three men weighing gold dust or a man committing
suicide belong to this group. There is no archaeological corpus that can match
the wealth of this ensemble. If objects are indeed the footsteps of human

regalia, imports, the

statuses, travellers, friends meeting, a

activity,

it

can

all

be reconstructed here.

At the same time the goldweights also underline the special traps set by
collections of objects and art works. We may be tempted to find that what was
201

ART HISTORY

IN AFRICA

Plate 11.2 Akan goldwexghis (after T. Garrard 1980; plaie 54). Brass cast by the lost wax technique. Late period
(1700-1900). The appearance of objects such as these boats, charms, ladder, key, treasure chests, sandals, claw, ring
and jug and for some, even their existence in the period, is knoum only by models made as goldwetghts. The exact date of

manufacture for each weight

202

is

unfortunauly not known. The date

'late period' is inferred

from

the style

ART IN HISTORY
not represented did not ekist, and that which was not frequently depicted was a
rare activity. Was suicide rare because only one surviving weight depicts it?
Perhaps. But was gold weighing rare because only one weight shows it?
Certainly not! Because compositions were limited in size (after all the objects
had to conform to a given weight!) there are no goldweights showing the grand
processions of rulers on days of national importance. There are thus limitations
to the information portrayed. Negative evidence should not count with this
type of source, especially with regard to assemblages.
Goldweights can be anachronistic, i.e. , represent an object or a scene long
after the object fell out of use or the activity had ceased. It survived in art by
replication. From Benin we know that even recent carving includes a frieze of
tiny Portuguese heads, replicated

from those found on sixteenth-centiuy

work.
Goldweights often come with commentary. They often represent
proverbs, making their meaning or symbolism expUcit (Appiah 1979; Menzel
1968). Again there are dangers here. The proverbs constitute a potential
storehouse of Akan practical philosophy, and of their cosmology, but we do
not know that they were in the thoughts of the maker of the weights who could
have had quite a different allusion or proverb in mind. Later the object itself
suggested other interpretations and other proverbs. Thus we can be almost
certain that the proverbs now cited along with the weights reflect
nineteenth-century interpretations, and nothing thsit certainly is earlier. Again,
anachronism is the main danger of interpretation accompanying the art work.
Modern interpretations must fully account for the stylistic conventions
current at the time. These are themselves historical evidence. Just as we do not
believe that Pharaohs were much taller than their subjects, we should not hold
that all Nubians or West Asians or Libyans correspond to the stereotyped
representations we find in Egyptian art. They are stereotypes and inform us

about Egyptian prejudices

at the

time they were created and that is all.

is that they may serve as the starting

A last general effect of works of art

point for oral elaborations, precisely when interpretative speculation grew

about them. Thus, Molet (1974) shows that several mythical animals entered
Malagasy folklore from the bottom of Chinese dishes exported to the island.
Some ceramics survived in deposits on the northwest coast and the dragons and
other phantasms drawn in porcelain are the prototypes of the strange and
sensational monsters in folklore, exactly as dragons are in our own folklore.
Iconatrophy, as this phenomenon is called, is also well known from elsewhere,
for example, in Ethiopia (Doresse 1972:215).

ART

IN

ECONOMIC AND SOCIAL HISTORY

objects testify to a technology. Representative art, such as the

goldweights, extends our knowledge of objects and relates directly to process
and production and consumption. Objects and art objects relate to trade in
other ways. Mediums were imported as the copper alloys were in Igbo Ukwu or
the brass in Benin, or foreign objects are mute witnesses to contact that must
have taken place. North African Mamluk copper vessels from fourteenth- or

Man-made

fifteenth-century Egypt in northern Ghana are evidence of contact (Posnansky

1979:53; Cole and Ross 1977). Styles can be evidence of contact either by the

203

ART HISTORY

IN AFRICA

presence of 'transitional' style or by foreign influences on local style. Thus,

some fifteenth- and again eighteenth-century Ethiopian paintings show
evident Far Eastern influences, perhaps Japanese (Leroy 1967:41; Chojnacki
1973:38). Architecture there, a century earlier, has Unks with Mogul India.We
know of the Indian connections via Goa from other sources; we did not know at
all about links with the Far East, however indirect they may have been. In most
cases such links were trade contacts, especially when several objects are found
together. Finally, art objects can tell us something about relative values of
materials. If ivory was reserved for the king, it was valuable; if it served to
make pestles to pound grain it was not. The first case was true in Benin, the
second in the villages along the Sankuru. If we say that bronze or brass was the
equivalent of gold in some societies, we base this on the restricted use of objects
made in such metals, the types of object (e.g., jewels) and the social context
inferred from the use (Herbert 1973).
In social history, art objects are as valuable a source as in economic
history. For they are art objects and not just utensils marked as something
special. They are tied to institutions. The skill displayed in them testifies to
degrees of specialization of labour. The distribution of their use relates to social
stratification. Above all, concentrations of art objects around cenain
institutions indicate their dominant role in society. If art is concentrated
around courts in states, around men's meeting houses in portions of Gabon and
Cameroon, around mosques, churches or such temples as exist among the
Tsogo of southern Gabon, around medicine and heahng as in eastern Kasai
(Zaire), or among several centres of attention, it tells us directly and
unambiguously what the important foci of society were. Lastly, some forms of
art, especially personal art, i.e., body art, or art typically linked to a person
rather than to a community, were commonly used to mark social strata,
emphasize status and role. Much of this evanescent art, such as scarifications,
hairstyles, differences in dress, has survived in painting or in sculpture.
Social historians simply do not yet realize the full potential of art in such
matters. Nor do they appreciate sufficiently that art is a direct source,
emanating from the community itself, not from foreign reports as is so often
the case in sub-Saharan Africa. Historians are now sensitive to archaeology as
evidence for social strata, for instance, as evidenced by differences in grave
goods, or as evidence for social coercion and control when e.g. 'imperial style'
buildings appear isolated from the main centres of rule in the Roman Empire
or in the Zimbabwe kingdom. They are still not directly interested in the
gruesome evidence for coercion in Ife and Owo as shown by the tied and
gagged sacrificial victims or the baskets of cut-off heads (ill. in Eyo and Willett
1980:41-2). They have not yet attempted (perhaps wisely, given the quarrels
over seriation!) to use Benin plaques and stylistic change in the royal heads as
evidence correlating to oral traditional data about the fluctuations of power in
Benin between the king and various groups of chiefs. And yet the changing
proportions in the representation of kings and others, or the increasing portion
paraphernaUa occupy on royal heads is testimony to relative positions and to
increased isolation of the king in a ritual position.
The potential of art objects to provide a better understanding of social
history is so great that we could align example after example of different
conclusions that can be drawn from such evidence and have not been, whether

204

ART

IN

HISTORY

be about the growth and spread of such institutions as associations, or about

changing social strata, or about the evolution of statuses or roles, or about
expressions of ethnicity. Only by using such sources can a full reaUzation of
their potential be acquired. But beyond their value, even as documents about
social history, works of art are unique as crucial data about ideology,
legitimation and world view.
it

ART AND INTELLECTUAL HISTORY

Sources for intellectual history in the portion of Africa where writing was
unknown or unused are skimpy indeed. We are in danger of writing history as
trends
if acting forces were driven merely by ecological, economic or social
little
know
and
we
past,
the
in
'see'
really
cannot
often
which we postulate, but
about aspirations, values, cosmologies, the role of individual thought or even
action. As an expression of culture art addresses these issues. Because of the
problems of interpretation it poses, art cannot be a panacea for all other gaps in
the record, but it certainly draws attention to this side of the past and sheds

some light on it.

Changing ideology,

legitimation and altering values can often be directly

documented. For instance, among the Songo of Angola a new sort of icon was
made in the second half of the last century. It showed a trader on an ox
accompanied by a bird, a white oxpecker, the symbol of luck and weahh and

masculinity (Bastin 1968/9: esp. 78, 80, Fig. 10). In works of a neighbouring
society and of the same period, among the Chokwe, the motif of a well-defined
mask and hat connotated power and wealth. With the expansion of the
long-distance trade, the notion of wealth gained at the expense of the other
elements that went into the make-up of power, and the motif appeared
prominently on Chokwe thrones. The Songo icon and the Chokwe motif speak
directly of

in values: wealth, obtained by trade brought prestige, a

competed with increasing success against the older prestige

change

prestige that

vested in the chieftaincy.

The comparison of synchronic representations in different cultures
vividly contrasts the meaning of roles that otherwise might seem similar. For
the Kuba, womanhood was a dramatic and sad role, best expressed by the
statue, ngaat apoong, set up in front of the initiation wall. She was seen

and as a wife. As a sister she was doomed to be taken away

from her near and dear to live among strangers. That was symbolized by her
being tied to a circle - at other times in a circle - of raffia at the foot of the wall,
just under the highest mask, that of her brother. The facial expression and
pregnancy also relate to this (Plate 11.3).
On Kuba masks of women, ngady amwaash, the same role was expressed
by the stylized tears. In contrast, the Chokwe in their masks and the Dan of
Liberia and the Ivory Coast in their masks and rare statues of women, glorified
female beauty and sensuous pleasure (Himmelheber 1960: 187). Elsewhere, as
among the Akan or on the coast of Zaire, motherhood was emphasized, even
though it was the motherhood of chiefs in a matrilineal society. They were the
originators of dynasties, the source of noble men. In southeastern Zaire,
femininity was obviously praised, but, despite their beauty, women were
portrayed as holders of ambiguous mystical powers (see Plate 5.5), as servants

explicitly as a sister

205

ART HISTORY

IN AFRICA

Plate 11.3 Female figure. Wood, white clay, painl, hair. Used
initiation.

Mapey, Kuba

country, Zaire.

boys'

Komnklijk Museum voor Midden

Afrika, Tervuren. Height 146-5cm. Photo 1953.

Made

shortly before(?)

ART

Plate 11.4 Charm. Wood.

Named

ishak

indweemy

such charms represent female nature

spirit.

IN

HISTORY

Exact use unknown.

University of Pennsylvania. Height 33cm. Acquired in Africa before 1920. Ishak indweemy
unknown. A statuette of similar style and size was acquired on the Kinshasa market tn 1971 from a

Kuba kingdom. Museum,

styles

remain

still

Kuba hawker

207

ART HISTORY

IN AFRICA

of men, or - on the contrary - as great ancestresses (Flam 197 1). In Nubia, the
Madonna stood for womanhood, and especially for the motherhood of the
supreme being (see Fig. 5.4). Of all mankind she had been the most important,
she interceded for men in heaven, and she towered over all men.
Interpretations of these works all date from the time of collection, usually
at the turn of the century, except for Nubia. They are all well documented,
based on a whole corpus of data and not contradicted by other icons, at least not
directly.

Other representations of women complement these. Thus, for instance,

and 8.5 show Kuba women in other roles: the pregnant daughter and

Plates 6.6

the sensuous provider of food. In addition,

Kuba female charm figures greatly

(Plate 11.4). But they do not

represent female nature spirits and the iconography,
as well as the style reflects this. Representations of women elsewhere among
the cases cited do not vary as dramatically, even though females are depicted in

differ

from the statues and the masks discussed

represent

women. They

other roles, such as the female saints in Nubia.

Our examples remain superficial in that we have not contrasted the full
range of meanings associated with womanhood in all these cases but only what
are seen to be the main ideals. Nevertheless, even this comparison allows us to
develop our sensitivity towards the variation in meanings associated with the
role of woman, meanings that are then to be further developed by examining
the full range of expression both in the visual and in the performing arts.
Legitimation and ideology are powerfully expressed in works of art
relating to rule. The ruler's enclosure as well as the minbar in mosques directly
link rule and religion and tell us how religion legitimized rule every Friday.
The representation of persons crucial in founding myths gives us a visual
legitimation of society, that allows us at the very least to see how that society
visualized its founding heroes at the time the works were made. Earlier we

discussed the statues of Chibinda Ilunga, the founding hero of the Chokwe,
statues that celebrated hunting, physical power, the various supports of the
supernatural. This is how Chokwe, relating the story of Chibinda Ilunga, saw
their hero in the nineteenth century and not the way we see him with our
imaginations now.
In a society along the upper Kwango, that of the Holo, there were statues
made of the 'Queen of the Holo' (Plate 11.5), and indeed there were queens
ruling there during the 1880s. But, more importantly, queens had been highly
important in that region since the seventeenth century when the heroine
Nzinga of Matamba fought the Portuguese and the Dutch and finally founded
the kingdom of Matamba west of the Kwango River, just across from the Holo
(Miller 1975). The statue of the queen reflects the Holo view of their

foundress, but probably also that of the ideal woman ruler, Nzinga. No
Westerner would recognize a queen in this statue of a woman seated on a
cylinder around which heavy iron wire was loosely twisted. Even a superficial
glance shows three remarkable characteristics: first, the sexlessness of the
person: no breasts, no sex. This was not a woman, it was a ruler. High status is
merely indicated by the coiffure and femininity only by the holes in the ears.
Second, the face differs from conventional Holo representations giving it a
strong expression of will and dynamic resolve rather than the usual serenity.

That makes
208

this the portrait of a ruler.

And

lastly there is the metal,

which

ART

IN

HISTORY

Plate 11.5 Siaiue of a queen of the Holo. Wood, iron, yellow beads and one red one. Use unknown. Upper middle
Collection J. Hautelei. Height 39cm. Date of acquisition unknown

Kwango, Zaire/Angola.

209

ART HISTORY

IN AFRICA

alludes to the special status of the Holo in the area as famed blacksmiths. The
lack of femininity and the ruthless will agree well with the written data existing
about Nzinga of Matamba and contrast totally with the representations of her
imagined by Europeans in the eighteenth century where she is very feminine
irresolute! (European illustrations: 17th c. Cavazzi 1965: vol. 2, 67, 68, 73;
Bassani 1977:xiii; 18th c. Davidson 1961:18.) The contrast could not be
greater. Thus the representations of ancestors or founders bring personality
back into African history and show us the visual images later generations had

and

about such personalities.

Art also expresses cosmology, sometimes explicitly as in Christian art or in
the sculpture of the Dogon, sometimes only by implication. Explicit

cosmology allows us easily to follow different threads that tie it to social

organization. Thus, the emperors of Ethiopia and its warrior saints were
almost interchangeable. In Nubia rulers protected by the Virgin were puny in
relation even to saints and angels. In Islam finer differences in mosques
emphasized local shades of theology. Thus, when the Almohads fought the
Almoravids, they legitimized their bid for power by the assertion that the
Almoravids deviated from Islam. Almohads stressed the 'oneness' of God
(their name means 'one'), the theology of the oneness of God (tawhid), and a
humble and mystical approach to God. Their mosques were severe and plain.
When they took Fes they whitewashed the rich decorative patterns in the
honeycombed domes of the aisle leading up to the mihrab (Terrasse 1968).
That expressed their legitimation of their bid for power. The power itself, they
expressed in massive military architecture, in which proportions rather than
decor were all important.
South of the Sahara, cosmology was rarely directly expressed, but the
conventions of art imply it. From the proportions it is clear that on the Benin
plaques kings towered over other men. The corpus of Kongo and related art of
lower Zaire stresses the real hidden world and the powers of charms, not of
gods. Thus, an ivory horn of unknown age from Loango depicts the founding
'Queen of the Vili' seated on her elephant recalling the foundation myth and its
cosmology. Hidden deep in the horn another statue has been detected by
X-rays. The ensemble renders a world view in iconic form, dissociating the
visible

and the

real

(ill.

Neyt 1981:89).

of art, regalia in kingdoms, sets of objects belonging to

overarching institutions elsewhere, are especially valuable both to document
legitimation and to provide insight into cosmology. Unfortunately, very few of
such sets could be studied until recently. As long as the institutions had deep
meaning, their icons were not accessible. Now such icons are beginning to be

Among works

studied.

What

they reveal

is

that cosmologies

and ideologies

alike

were

far

condensation of
cosmology, theories of social organization, legitimations of existing order
and pose obvious problems of interpretation. Much of the local interpretations
about these objects will be of recent vintage, because the original meaning of
many features has been overlaid. Nevertheless the outlines of older meanings

more complex than

is

usually

imagined.

They

are

should

210

still

be discernible.

ART

IN

HISTORY

THE RELEVANCE OF ART TO HISTORY

Ultimately art history is relevant to general history because of two special
properties: the concrete character of works of art and the fact that trends in art
directly

document meaningful change

The houses
character of

art.

of

Art

generalization and

it

Lamu

at large.

in the eighteenth century illustrate the concrete

is the past coming to us without simplification, without

comes to us at a glance. It is an ideal mode of expression to

render a situation directly rather than to describe it with all the selectivity
description entails. This opaque character of art, this power of confrontation,

who is not used to being brought face to face with a

foreign reality from the past. Art works do not narrate, they do not argue. They
present a whole at once. This property implies a complexity of everything in
baffles the historian,

use, social, emotional and mental significance, a

were. To explore and to understand a work of art fully
one must relate it to everything that is known about the period and the place
from where it stems. Even so, it almost always raises new problems to be solved
and does not just confirm the already known. Its major challenge to historians
though is that art acts as a total proof. After an historian has made all his
dissections of data, traced all his causalities, built up his explanatory theory, it
must stand the test of art. It should explain everything about this testimony to
the 'spirit of the age'. The object from the past should fit naturally in the

one:

medium, technology,

frozen slice of life as

it

it is because of this power to challenge, to

uncover precisely what remains unknown and what is known, that art works
are important to historians.
Trends in art history should also reflect social and cultural history, for
they are an expression of that history both in its superficial alterations and in its
deep, often barely conscious changes. A few examples make this clear. We
start with a clear-cut and simple case of culture contact. A drawing was made
on the door of the house of the headman of the village of Lemba in 1885 or
shortly before (Plate 11.6). It is a mixture of local style and European
influences. Lemba was then some ten miles away from Leopold ville where the
European post had only been established in December 1881. Barely three to
four years later that fact was reflected not only in a passing sketch but as a
received alteration in style consecrated by the eminent place in which it was
displayed. It was taken away from Lemba during a raid by Europeans in
December 1885 or January 1886(BulaNzau 1894:194-7; Krieger 1969: vol. 3,

reconstruction of the past, and

56, Bild 180; Biographic Coloniale Beige 1951: vol. 2, 690-1). Art works
indeed reflect events!
Art expresses changes in mentality of long duration. Thus we find that
before the advent of the Almohad dynasty, Moroccans had already begun to
elaborate a cult for the dead, something that, strictly speaking, is outlawed in
Islam. They built an addition to the main mosque at Fes on the outside of the
mihrab wall, near the focal point of the whole oratory, so that corpses could be
put near the mihrab without being in the mosque, which they would have
polluted. A little over a century later, in 1310, the Marinids laid out the first
dynastic cemetery with an oratory and mausolea at Chella, while in the
countryside, well-built domed tombs began to be erected for renowned saintly
men. In very little time such tombs came to be surrounded by a mosque and a

211

ART HISTORY

IN AFRICA

Plate 11.6 Doorpanel. Wood, raffia stems, vines, paint black and white. Used in the house of the headman of Lemba
Painted
village. Hum people, Zaire (near Kinshasa). Museum fiir Volkerkunde, Berlin-Dahlem. Height 84-3cm.

1882-1885

school as the descendants and disciples of the holy men gathered and as the
tombs became centres for pilgrimage. By the 1350s, brotherhoods were
formed, founded by mystical holy men, whose lodges formed around the tomb
sites of saints. Such zawiya were soon to become such a force in the land that all

Morocco have had to cope with them. The architectural

evidence shows this spread of the cult for the dead, for saints in particular, and
correctly shows that the trend was older than the Marinid dynasty. The
founding saints of Moroccan Sufism lived earlier under the Almohads,
themselves mystically inclined. But as the mosque at Fes shows, the cult of the
future governments in

dead had begun to take hold even before the Almohads were well established,
212

ART

IN

HISTORY

even before Sufi saints appeared. Despite their purism the Aimohads could not
arrest this development. It was only deflected and became a cult for the saints
only. Art here testifies to a long and major shift in Moroccan society and in
Moroccan mentality as it developed over two centuries (Terrasse 1932, 1968;
Margais 1954:281-4, 299-301; Laroui 1977:245-6).
More general and longer trends can also be correlated. Coptic art as a
whole shows increasing tendencies towards stylization, away from the extreme
naturalism which was its milieu of origin and became a highly ascetic art (du
Bourguet 1967; Wessel 1965; Badawy 1978). In an evolution spanning well
over five centuries the trend was never reversed. We can link it to the facts that
Coptic art was Christian art above all and that it was an art for the dispossessed
whether by and for monks, away from the Nile valley, or for the middle class

and the poor in the valley itself. The evolution was a spiritualization of art,
where meaning rather than full figuration counted for more and more. It is
thus possible to account for the evolution of art, of social and of intellectual
history at once. Nevertheless, difficulties remain, for monks did not become
ever more ascetic with the passage of time, nor did the common people become
more dispossessed, nor were all Copts always poor. Rather, the trend
continued under its own momentum both in art and in religious expression.
Moreover, even before the Arab conquest, stylization had become a
profoundly perceived ethnic trend. Coptic religion and Coptic art were typical
Egyptian features.
With this last example we reach two limits. The reader will appreciate that
when we deal with such time spans and such complex correlations, our sketch
becomes less and less convincing for lack of detailed evidence which requires
more space than I have in this book. But we also reach a limit of what can be
proved. If we went beyond this limit we could, for instance, argue that the
development of geometrical arabesques to new heights and new mechanical
precision of execution under the Marinids is to be related to the increased role
of lawyers in public life and of law as a way of thinking with its sharp concepts
and its long consequences. But the link is farfetched. This also was an age of
rising mysticism. So why not correlate the conquest of voids by geometry with
mysticism? Such links cannot be proved, merely suggested. Because they
cannot be demonstrated by detailed correspondence they remain fruitless. But
that last case does illustrate the particular danger of using art history within a
general framework: namely, to let imagination run away with what become
personal interpretations of an age.
Art history is important to general history. It brings sources, it brings us
an immediacy from the past that cannot be replaced or ignored, and it gives us a
further tool to probe the significance of both small breezes and long trade
winds in the atmosphere of history. Art is a weathervane of history, and more.
Art, produced by forces outside itself, expresses metaphors which in turn can
lead to further cultural and social change.

213

REFERENCES AND
FURTHER READING
BIBLIOGRAPHIES
BASKIN,

L.

P. (1968)

J.

An

Annotated Bibliography

on Ethiopian Art, Addis

Ababa

BURT,

E. C. (1980)

An

Annotated Bibliography of the Visual Arts of East Africa,

Bloomington, Indiana

COULET WESTERN,

D. (1975) Bibliography of

the Arts

of Africa, Waltham,

Massachusetts

CRESWELL,

K. A. C. (1961)

Bibliography of the Architecture, Arts and Crafts of

Islam, Cairo

GASKIN,

J. P.

(1965)

/I

Bibliography of African Art,

KAMMERER, W.

(1950)

Coptic Bibliography,

L.

OYENIYI, OSUNDINA

Ann

London

Arbor, Michigan

(1968) Bibliography of Nigenan Sculpture, Lagos

JOURNAL ABBREVIATIONS
AA:
A AN:
AT:
J

AH:

African Arts
Arts d'Afrique Noire

Africa-Tervuren

Journal of African History

REFERENCES
(1977) Nubia, Corridor to Africa, Princeton, New Jersey
(1972) History of the Niger Delta, Ibadan
APPIAH, P. (1979) 'Akan symbolism', AA, 13 (1), 64-7
ARKELL, A. J. (1950) 'Gold Coast copies of 5th-7th centuries bronze lamps',

ADAMS, W. Y.
ALAGOA, E. J.
Antiquity, 24,

ATHERTON,

J.

38^0
and

KALOUS, M.

(1970) 'Nomoli,

J AH,

11 (3),

303-17

ATIL,

E. (1981) Art of the Mamluks, Washington

AYOUB, A. and GALLEY, M. (1977) Images de Djazya, Paris.

BADAWY, A.

{\91%) Coptic Architecture and Archaeology, Cambridge, Massachusetts

(1968) 'Kitab almasalik wa'l mamalik (1067/8)'. Translated by V. Monteil,
'Routier de I'Afrique blanche et noire du nord-ouest'. Bulletin de Vlnstitut

al-BAKRI

F ondamental d'Afrique Noire B, 30, 39-116

BALANDIER, G. (1968) Daily Life in the Kingdom of Kongo, from
Eighteenth Century, New York
214

the Sixteenth to the

REFERENCE MATERIAL

BASCOM, W.

(1973) African Art

in

Cultural Perspective,

New York

BASSANI, E. (1977) Scultura Afncana Nei Musei Italiam, Bologna

BASSANI, E. (1978) 'Les sculptures Vallisnieri', AT, 24 (1), 15-22
BASTIN, M. L. (1965) 'Tshibinda Uunga' Baessler Archiv, NF 13, 501-37
,

BASTIN,M.L.

(1968/9) 'L'artd'unpeupled'Angolal.Chokwe

.3.

Songo',AA,2

(2), 40-6, 60-4; (3), 50-7, 77-81

BASTIN, M. L. (1976) 'line statuette de Tshibinda Ilunga (Tshokwe, Angola)
disparue d'un musee portugais (ou elle se trouvait depuis 1914) mutilee dans un but

defraude)',Ar, 22(1), 4-8

BATTUTA,

(Ibn) (1966) Tuhfat al-nuzzar fi ghara'ib al-amsar wa-'aja'ib al-asfar

(1352/3 voyage). Trans. V. Monteil, R. Mauny, A. Djenidi, S. Robert, J. Devisse,
Extraits Tire des Voyages, Dakar
H. (1929) Afrikanisches Kunstgewerbe in Th. Bossert (ed.), Geschichte des
Kunstgewerbes, 6 vols. Berlin, vol. 2, pp. 104 seq.
H. (1964) 'Die Ethnologische Beurteilung einer Vorgeschichtlicher

BAUMANN,

BAUMANN,
Keramik

Mittelafrika'

in

in

Festschrift fUr

A. Jensen, Munich,

vol.

1,

pp.

13-58

H. (1969) Afnkanisc he Plastik und Sakrales Konigtum, Munich

R. (1977) Tellem:een Bijdrage tot de geschiedenis van de Republik Mali,
Utrecht (Tellem: a contribution to the history of the Republic Mali)
BEDAUX, R. (1980) 'The geographic distribution of footed bowls in the Upper and
Middle Niger region' in B. K. Swartz and R. E. Dummett (eds). West African

BAUMANN,
BEDAUX,

Culture Dynamics,

The Hague,

pp. 247-58

BEDAUX, R., CONSTANDSE-WESTERMANN, T. S., HACQUEBORD, L.,

LANGE, A. G., VAN DER WAALS, J. D. (1980) 'Recherches archeologiques
dans

le

Delta interieur du Niger', Palaeohistoria (1978), 91-220

BEN AMOS,
BEN AMOS,
BEN AMOS,

P. (1976)

P. (1980b) 'Patron-artist interactions in Africa,'

BERNATZIK,
4,

in Benin Art', Man, (NS) 11, 242-52

The Art of Benin, London

'Men and Animals

P. (1980a)

AA,

13 (3), 46-7, 92

H. A. (1932) 'Meine Expedition nach Portugiesisch-Guinea', Ar/anris,

197-211

BERNATZIK, H. A. {\9ll) Aethiopien des Westens, Vienna

BIEBUYCK, D. {1911) Lega Culture, Berkeley, California
Biographie Coloniale Beige (1948-1977) Brussels, 7 vols.
BIRMINGHAM, D. and MARTIN, P. (eds) {\9%l)History of Central Afnca, 2 vols.,

London
BISSON, M.

(1975) 'Copper currency in Central Africa -the archaeological evidence',

World Archaeology, 6 (3), 276-92
BONTINCK, F. (1979) 'La provenance des sculptures Vallisnieri', AT, 25 (4), 88-91
BRAIN, R. (1980) Art and Society in Africa, London
BRANDENBURG, D. (1966) Islamische Baukunst in Aegypten, Berlin
BRAVMANN, R. (1973) Open Frontiers, New York
BRAVMANN, R. (1974) Islam and Tribal Art in West Africa, Cambridge
BREUIL H. (1955) The White Lady of the Brandberg, Paris and London
BROWN, J. L. (1980) 'Miji Kenda Grave and Memorial Sculpture', AA, 13 (4), 36-9,

88

G. (1910) 'Les populations de la Moyenne Sanga', Revue d'Ethnographte et de

3-32 (Porno and Bomwali)
NZAU (BAILEY), H. (1894) Travels and Adventures in the Congo Free State and

BRUEL,

Sociologie, 1,

BULA

its Big Game Shooting, London

Bundesministerium fiir Unterricht (1970) Nubische Kunst aus Faras, Vienna

THE CAMBRIDGE HISTORY OF AFRICA (1975-78) R.

vols.

Oliver (ed.), Cambridge,

2-5

215

ART HISTORY

IN AFRICA

CAVAZZI DE MONTECUCCOLO,

G. A. (1965) Istorica Descrizione del tre regni:

Angola, Bologna 1687, trans. G. M. De Leguzzano, Descricdo
historica dos Tres Reinos: Congo, Matamba e Angola, 2 vols, Lisbon (Historical
Description of the Three Kingdoms: Congo, Matamba and Angola)
CHAFFIN, A. F. (1979) L'Art Kota. Les figures de reliquaire, Meudon
CHITTICK, N. (1974) 'Excavations at Axum: a preliminary report', Azania, 9,

Congo,

Matamba

195-205

CHITTICK, N.

CHOJNACKI,

(1976) 'Radiocarbon Dates from Axum', Azania, 9, 179-81

(1973) 'Ethiopian Paintings', Religious Art of Ethiopia,

S.

n.l.

(Recklinghausen?)

CHOJNACKI,

(1977) 'Two Ethiopian Icons', AA, 10 (4), 44-7, 56-61, 86

A. (197 \) Afrikaanse Kunst, Utrecht (African Art)
A. (1978) 'Een ongewone edan-ogboni', De Vrienden van het
Etnografisch Museum Antwerpen, 5 (2), 2-3 (An unusual edan-ogboni)
COART, E. (1927) Les Nattes, Brussels
COART, E. and HAULEVILLE, A. (1907) La Ceramique, Brussels
COLE, H. (1969a) 'Mbari is life', AA, 2 (3), 8-17, 87
COLE, H. (1969b) 'Mbari is a dance', AA, 2 (4), 42-51, 79
COLE, H. (1969c) 'An is a verb in Iboland', AA, 3 (1), 34-9
COLE, H. (1975) 'The history of Ibo mbari houses: facts and theories' in D. F. McCall
S.

CLAERHOUT,
CLAERHOUT,

New York, pp. 104-32

(1977) The Arts of Ghana, Los Angeles
D. (1973) 'Throwing knives in Equatorial Africa: a distributional study',

and E. Bay

(eds), African Images,

COLE, H. and ROSS, D.

CORDELL,

Ba-Shiru, 5

CORNET,
CORNET,
CORNET,
CORNET,

(1),

94-104

(1972) Art de t'Afrique noire au pays du fleuve Zaire, Brussels

J.

Am

of Africa. Treasures from the Congo, New York

(1976) 'A propos des statues ndengese', AAN, 17, 6-16
(1978) A Survey ofZainan Art: The Bronson Collection, Raleigh, North
(1975)

J.
J.
J.

Carolina

CORNET, J. (1980) 'Pictographies woyo', Quademi Poro, no. 2, Milan

CRESWELL, K. A. C. (1952-9) The Muslim Art of Egypt, 2 vols, Oxford
CRINE-MAVAR, B. (1968) 'A propos de la sterihte artistique des Lunda',

Etudes

Congolaises, 11 (2), 59-67

CROWE,

D. W. (1971) 'The geometry of African

Geometry,

DANIELS,

DARK,
DARK,

1,

1.

Bakuba

An\ Journal

of

C. (1970) The Garamantes of Northern Libya, Stoughton, Wisconsin

P. J. C. (1973) 'Brass casting in West Africa', AA, 6 (4), 50-3, 94

P. J. C. (1975) 'Benin bronze heads: styles and chronologies' in D. F.

and E. Bay

DARK,

art:

169-82

(eds), African Images,

P. J. C. (1981)

New

'Head and Tusk'

in

McCall
York, pp. 25-103
S. Vogel (ed.), For Spirits and Kings, New

York, 136-40

DAVIDSON,

B. (\96\) Black Mother, Harmondsworth and Boston, Massachusetts

O. (1967) West Africa Before the Europeans, London
d'AZEVEDO, W. L. (1973a) 'Sources of Gola Artistry' in W. L. d'Azevedo (ed.). The
Traditional Artist in African Societies, Bloomington, Indiana, pp. 282-340
d'AZEVEDO, W. L. (1973b) (ed.) The Traditional Artist in African Societies,
Bloomington, Indiana
DECHAMPS, R. (1970/8) 'L'identification anatomique des bois utihses pour des
sculptures en Afrique', AT, 16 and 24
DE GRUNNE, B. (1980) Anctenr Terra Coltas From West Africa, Louvain-la-neuve
DE HEUSCH, L. (1956) 'Le symbolisme du forgeron en Afrique', Reflets du Monde,
10, 57-70

DA VIES,

216

REFERENCE MATERIAL

DELANGE, J. (1974) The Art and Peoples of Black Africa, New

DE MARET, P. (1977) 'Sanga: new excavations, more data
problems', JA//, 18 (3), 321-38
DE MARET, P. (1980a) 'Ceux qui jouent avec

le feu: la

place

York
and some related

du forgeron en Afrique

Afnca, 50 (3), 263-79

MARET, P. (1980b) 'Les trop fameux pots a fossette basale du Kas^i\ AT, 26

centrale',

DE

(1),

4-12
L. (1980) Techniques et Economie des Pygmees Babinga, Pans
E. (1971) 'Use of paper chromatography to determine the age of
albuminous binders and its application to rock paintings', South African Journal of

DEMESSE,

DENNINGER,
Science, 2,

80-4

DENVER, S. (197 S) African Traditional Architecture, London, New York

DE SOUSBERGHE, L. (1959) L'Art Pende, Brussels
DE VERE ALLEN, J. (1974a) 'SwahiU architecture in the later Middle Ages', AA, 7
(2), 42-7, 66-8, 83^
DE VERE ALLEN, J. (1974b) 'Swahih culture reconsidered: some historical
implications of the material culture of the northern

Kenya coast in the 18th and

19th

centuries', Azania, 9, 105-38

DE VERE ALLEN,

J.

and

WILSON,

T. H. (1979) Swahili Houses and Tombs of the

Coast of Kenya, London

DINKLER, E. (1970) Kunst

und Geschichte Nubiens in Christlicher Zeit,

Recklinghausen
DITTMER, K. (1967) 'Bedeutung, Datierung und Kulturhistorische Zusam-

menhange der "prahistorischen"

Baessler Archiv,

NF

15,

Steinfiguren aus Sierra

Leone und Guinee',

183-238

before the 15th century, in J. Ki Zerbo (ed.),

J. (1980) "Written sources
General History of Afnca, London, vol. 1, pp. 87-113
DORESSE, J. (1972) La Vie Quotidienne des Ethiopiens Chretiens (au XF//' et au

DJAIT, H.

XVnr

siecles),

DREWAL,

Paris

H. (1980) African

Artistry: technique

and

aesthetics in

Yoruba

sculpture,

Atlanta, Georgia

H. (1981) 'Staff (Edan Oshogbo)' in S. Vogel (ed.). For Spirits and Kings,
York, pp. 90-1
DU BOURGUET, M. P. (1967) The Art of The Copts, New York
DUPRE, M. C. (1968) 'A propos d'un masque teke de I'Ouest Congo', Objets et
Mondes, 8, (4);, 295-310
DUPRE, M. C. (1979) 'A propos du masque teke de la collection Barbicr' ,Connaissance
des Arts Tribaux. Bulletin du Muse Barbier-Muller, no. 2
DUPRE, M. C. (1980) 'I'art kota est-il vraiment kota?', I'Ethnographie, 83, 343-55

DREWAL,
New

E. and FAGG, W. (1958) The Sculpture of Africa, New York

Exoticophylacium Weickmannianum (1659), Anon., Uhn (Weickmann Catalogue)
EYO, E. (1978) 'Ekpe costume of the Cross Riwcf, AA, 12 (1), 73-5, 108

ELISOFON,
EYO,

E. and

WILLETT,

F. (1980) Treasures of Ancient Nigeria,

New York

in R. Oliver (ed.). The Cambridge History

J. (1977) 'Upper and Lower Guinea'
of Afnca, Cambridge, vol. 3, pp. 463-518
FAGG, W. (1959) Afro-Portuguese Ivories, London
FAGG, W. (1965) Tnbes and Forms in African An, New York
FAGG, W. (1970) Divine Kingship in Afnca, London
FAGG, W. and PICTON, J. (1970) The Potter's An in Afnca, London
FAIRLEY, N. J. (1978) 'Mianda yaBen'Ekie', Ph.D. theses. University of New York,

FACE,

Stony Brook, English text (Matters of the Ben'Ekie)

217

ART HISTORY IN AFRICA

FERNANDEZ, J. W.

(1971) 'Principles of opposition and vitality in Fang aesthetics'

and Aesthetics in Primitive Societies: a critical anthology,

in C. J. Jopling (ed.), Art

New

York, pp. 356-73

FERNANDEZ,
FERNANDEZ,

J.

J.

W.
W.

(1974) Review, AA, 8 (1), 76-7

(1975) Letter, A.4, 8 (2), 2

FISCHER, E. (1978) 'Dan Forest spirits: masks in Dan villages', /lA,

FLAM, J. (1971) 'The symbohc structure of Baluba caryatid

FLEMING, S. (1977) Dating in Archaeology, New York
FLEURIOT DE LANGLE, A. (1876) 'Croisieres a la cote
Monde, 31

(1),

1 1

stools',

(2),

AA,

d'Afrique',

16-23, 94
(2), 54-9

Le Tour du

289-304

FOCILLON, H. (1943) La Vie des Formes, Paris

FRANCASTEL, P. (1951) Peinture et Societe, Paris
ERASER, D. (1972) 'The fish-legged figure in Benin and Yoruba art' in D.

Eraser and

Art and Leadership, Madison, Wisconsin, pp. 261-74

ERASER, D. (1980) Review, AA, 14 (1), 15-22
ERASER, D. (1981a) 'A propos', AA, 14 (2), 27-8
ERASER, D. (1981b) 'Pair of armlets' in S. Vogel (ed.). For Spirits and Kings, New
York, pp. 128-9
FROBENIUS, L. (1896) 'Die Kunst der Naturvolker', Westermanns Monatshefte,

H. Cole

(eds), African

79-329-340, 593-606

FROBENIUS,

L. (1898) Masken und Geheimbunde Afrikas, Leipzig, vol. 79, pp.

329-40, 593-606
FROBENIUS L. (1907) Im Schatten des Kongostaates, Berlin
FROBENIUS L. (1933) Kulturgeschichte Afrikas, Zurich

R. (1969) African Crafts and Craftsmen, New York

GARLAKE, P. S. (1973) Great Zimbabwe, New York
GARRARD, T. (1979) 'Akan metal arts', AA, 13 (1), 36-43, 100
GARRARD, T. (\9S0) Akan Weights and the Gold Trade, London
GAULME, F. (1981) Le Pays de Cama, Paris
GEBAUER, P. (1979) Arr ofCameroun, New York
GERBRANDS, A. A. (1957) Art as an Element of Culture Especially

GARDI,

in

Negro Africa,

Leiden

GERSTER,

G. (1968) L'Art Ethiopien: eglises rupestres, Paris

U. (1973) 'Swahih plasterwork', AA, 6 (2), 46-59
GOLLNHOFER, O. and SILLANS, R. (1963) 'Recherches sur le mysticisme des
Mitsogo' in Reincarnation et vie mystique en Afrique noire (colloque de Strasbourg),
Strasburg, pp. 142-74
GOMBRICH, E. H. (I960) Art and Illusion, New York

GHAIDAN,

GOODY, J. (ed.) (1968) Literacy

GRAEBNER, F. (1911) Methode

in

Traditional Societies,

Cambridge

der Ethnologic, Heidelberg

GRIAULE, M. (194S) Dieu d'Eau, Paris

GRIGORIEVA, M. (19S0) Hermitage, Leningrad,

Leningrad

GROTTANELLI, V. (1975) 'Discovery of a masterpiece', AA, 8 (4), 14-23, 83

GRUNER, D. (1977) 'Der Traditionelle Moscheebau am Mittleren Niger', Paideuma,
23,

101^0

HALDANE,

D. (1978) Mamluk Painting, Warminster

H ALLEN, B. (1979) 'The an historian as conceptual analyst ',7ouma/ of Aesthetics and

Art Criticism, 37, 303-13
E. and

HAMMERSCHMIDT,

JAGER, O.

(1968) Illuminierte Aethiopische Hss,

Wiesbaden

HARTER,
218

P. (1972) 'Les

masques

dit

Bacham',

AAN,

3,

18-45

REFERENCE MATERIAL

HARTWIG, G. (1978) 'Sculpture in East Africa', AA, 11

HAU, K. (1959) 'Evidence of the use of pre-Portuguese

(4), 62-5, 96
written characters by the

Bini?', Bulletin de I'lnstitut Frangais de VAfrique Noire, B.21,

109-54

HAU,

K. (1964) 'A royal title on a palace tusk from Benin', Bulletin de I'lnstitut
Fondamental de VAfrique Noire, B.26, 21-39
HERBERT, E. (1973) 'Aspects of the use of copper in pre-colonial West Africa', J^A//,
14 (2), 179-94
HERREMAN, F. (1978) 'Doorlichting van een kafigeledio', De Vrienden van het
Etnografisch Museum Antwerpen, 5 (3), 2-A (X-ray: kafigeledio)
HILL, D. and GOLVIN, L. (1976) Islamic Architecture in North Africa, Hamden,
Connecticut

HIMMELHEBER,
HOAG,

J.

H. (1960) Negerkunst und Negerkunstler, Braunschweig

D. (1963) Western Islamic Architecture, New York

HODGEN, M.

(1942) 'Geographical distribution as a criterion of age', American

Anthropologist, 44, 345-68

HOLY, L. (1967) The Art ofAfrica. Masks and Figures from Eastern and Southern Africa,
Prague

HORTON, R. (\965a) Kalahari Sculpture, Lagos

HORTON, R. (1965b) 'A note on recent finds of brasswork in the Niger delta', Odu,
(1),

76-91

HRBEK,

L. (1977) 'Egypt, Nubia and the Eastern Desert' in R. Oliver (ed.),

Cambridge History of Africa, Cambridge, vol. 3, 10-97
HRBEK, L. (1980) 'Written Sources from the 15th century onwards' in J. Ki Zerbo
(ed.). General History of Africa, London, vol. 1, 114-41
HULSTAERT, G. (1931) 'Over de volksstammen der Lomela', Congo, 1, 13-52
(About the peoples of the Lomela)

R. R. (1979) The Peopling of Southern Africa, New York

'
E. (1968) The Ethiopian Church, Boston, Massachusetts
including dating
techniques
ISKANDER, Z. (1980) 'African archaeology and its
techniques' in J. Ki Zerbo (ed.). General History of Africa, London, vol. 1, 206-32

INSKEEP,
ISAAC,

L. (1980) 'The White Lady of the Brandberg. A re-interpretation',

Namibiana, 2(1), 21-9
JAEGER, O. A. (1965) Antiquities of North Ethiopia, Stuttgart
JOHNSON, M. (1977) 'Cloth strips and history', West African Journal of Archaeology,
7, 169-78

JACOBSON,

KAUENHOVEN JANZEN,

R. (1981) 'Chokwe thrones', AA, 14 (3), 69-74, 92

(1980) 'African Art at the Staatliches Museum fur Volkerkunde,
Munich', AA, 14 (1), 32^1
KI ZERBO, J. (ed.) (1980) General History of Africa, London, vol. 1, Methodology

KECSKESI, M.

and African Prehistory

KRIEGER, K. and KUTSCHER, G. (1960) Westafnkanische Masken, Berlin

KRIEGER, K. (1965-69) West Afnkamsche Plastik, Berlin, vol. 1 1965, vols 2

and

1969

KROEBER,
KROEBER,

A. (\9A%) Anthropology, New York

A. (1953) Cultural and Natural Areas of Native America, Berkeley,

California

A. and VAN LEYNSEELE, P. (1978) 'Social Anthropology and the "Bantu

Expansion"', Africa, 48 (4), 342-56

KUPER,

LABURTHE-TOLRA,

P. (1977)

Minlaaba, 3 vols, Lille

219

ART HISTORY

IN AFRICA

LAJOUX, J-D. (1977) Tassili n'Ajjer, Paris

LAMB, V. (1975) West African Weaving, London
LAROUI, A. (1977) The History of the Maghreb, Princeton, New

Jersey (Trans. R.

Manheim)

LAVACHERY, H. (1954) Statuaire de I'Afnque Noire, Brussels

LAWAL, B. (1977) 'The present state of art historical research in
(2),

Nigeria', JA//, 18

193-216

LAYTON,

R. (1981) The Anthropology of Art,

R. (1953) Les Bamileke, Paris

LECOCQ,

New York

DELANGE, J. (1967) Afnque Noire: la creation plastique, Paris

H. (1948) Sculptures Soudanaises, Paris
Late Middle Ages and During the Gondar
J. (1967) Ethiopian Painting in the

LEIRIS, M. and

LEM, F.
LEROY,

New York

Dynasty,

LEROY,

J.

(1968) 'La peinture ethiopienne au

Ethiopien, Paris, pp. 61-8

LEROY, J. (1973) L'Ethiopie: archeologie

LEUZINGER,

Moyen Age'

et culture,

in G. Gerster (ed.), L'art

Paris

E. (1980) The Art of Africa, Baden Baden (Trans, from the

German by

A. Keep)
(1973) Ancient Ghana and Mali, London
A. (1966) Architecture de I'lfriqiya, recherches sur

LEVTZION, N.
LEZINE,

les

monuments aghlabides,

Paris

LOIR, H.

(1935) Le Tissage du Raphia au Congo Beige, Brussels

McCALL, D.
New York
McINTOSH,

F. and

BAY,

E. (1975) African Images: essays in African iconology.

R. and McINTOSH, 5.. (1981) 'Empire of Mali: evidence from

Jenne-Jenno',7A//, 22 (1), 1-22
McNAUGHTON, P. (1970) 'The throwing knife in African history, AA, 3 (2), 54-60
MAES, J. (1938) Kabila en Grafbeelden uit Kongo, Brussels (Kabila statues and other
funerary statues from Congo)
MAGGS, T. and DAVISON, P. (1981) 'The Lydenburg heads and the earliest African
sculpture south of the Equator', AA, 14 (2), 28-33, 88
MAQUET, (1979) 'Art by metamorphosis', A A, 12 (4), 32-7, 90-1
J.

MAR^AIS,

G. (1954) L' Architecture Musulmane d'Occident, Paris

The External Trade of the Loango Coast: IS76-I580, Oxford
MAUNY, R. (1970) Les siecles Obscurs de I'Afrique Noire: histoire et archeologie, Paris
MEKHITARIAN, A. (1954) La Peinture Egyptienne, Geneva
MENZEL, B. (1968) Goldgewichte aus Ghana, Berlin
MENZEL, B. (1972) Textilien aus Westafnka, 3 vols, Berlin
MERCIER, M. (1928) 'Notes sur une architecture berbere saharienne', Hesperis, 8,

MARTIN,

P. (1972)

413-29
A. (1973) 'The Bala musician' in W. D'Azevedo (ed.). The Traditional
Bloomington, Indiana, pp. 250-81
MICHALOWSKI, K. (1967) Earas, die Kathedrale aus dem Wiistensand, Einsiedeln
MILLER, J C. ( 1 975) 'Nzinga of Matamba in a new perspective' AH, 1 6 (2), 20 1- 16
MOLET, L. (1974) 'Origine chinoise possible de quelques animaux fantastiques de
Madagascar', Jowma/ de la Societe des Africanistes, 64 (2), 123-38

MERRIAM,

Artist in African Societies,

MONNERET DE VILLARD,
I'illustrazione dei Miracoli di

U. (1947) 'La Madonna di S. Maria Maggiore e

Maria in Abissinia', Annali Lateranensi, 11, 9-90

Rome

MONTI DELLA CORTE,

A. A. (\94Qi) Lalibela,

MORTON WILLIAMS, P.

(1960) 'The Yoruba Ogboni cult in Oyo', Africa, 30 (4),

362-74

220

REFERENCE MATERIAL
R. H. (1904) Fetichism in West Africa, London
(1960) 'Quelques poteries protohistoriques a face humaine trouvees au
Katanga (CongoY, Journal de la Societe des Afncanistes, 30 (2), 145-50
NEYT, F. (1977) La Grande Statuaire Hemba du Zaire, Louvain-la-neuve

NASSAU,

NENQUIN, J.

NEYT, F. (1979) L'art Eket, Paris

NEYT, F. (1981) Traditional Art and History of Zaire,
NICKLIN, K. (1980) Review, AA, 13 (2), 24-7, 83

Brussels

OHNUKI TIERNEY,

perception/cognition/symbol-

E. (1981) 'Phases in

human

classification', American
451-67
OJO, G. J. A. (1966) Yoruba Palaces, London
OLBRECHTS, F. (1941) Bijdrage tot de Kennis van de Chronologie der Afrikaansche
Plastiek, Brussels (Contribution to the knowledge of the chronology of African
sculpture and graphic arts)
OLBRECHTS, F. (1943) 'Contribution to the study of the chronology of African
plastic art', Africa, 14 (2), 183-93
OLBRECHTS, F. (1959) Les Arts Plastiques du Congo Beige, Brussels (Trans, from the
Dutch by A. Gilles de Pelichy)
OLIVER, R. and FAGAN, B. M. (1975) Africa in the Iron Age, Cambridge

ization processes: cognitive anthropology

and symbohc

Ethnologist, 8 (2),

D. (1981) 'Figure pomdo' in S. Vogel (ed.). For Spirits and Kings, New
York, 61-2
PAULME, D. and BROSSE, J. (1956) Parures Africames, Paris
PECHUEL LOESCHE, (Dr) (1907) Volkskunde von Loango, Stuttgart
PERROIS, L. (1972) La Statuaire Fang, Paris
PERROIS, L. (1979) Arts du Gabon, Arnouville
PERSON, Y. (1961) 'Les Kissi et leurs statuettes en pierre dans le cadre de I'histoire
ouest-africaine'. Bulletin de Vlnstitut Fondamental d'Afrique Noire, B.23, 1-59
PHILLIPSON, D. W. (\911)The Later Prehistory of Eastern and Southern Africa, New

PAULME,

York

PICTON, Y. and MACK, Y. (1979) African Textiles, London

PISKATY, K. (1957) '1st das Pygmaenwerk von Henri Trilies

eine zuverlassige

Quelle?', Anthropos, 52, 33-48

POLAKOFF, C. (1980) Into Indigo: African textiles and dyeing techniques. New York
POSNANSKY, M. (1979) 'Dating Ghana's earliest art', AA, 13 (1), 52-3, 100
word', AA, 14 (4), 1-11
(1975) 'Perseus and Medusa in Africa. Military art in Fanteland:
1834-1872', AA, 8 (3), 36-41, 68-71, 91
PREYS, G. (n.d.) 'Dossier Ethnographique no. 427', Royal Museum for Central

POVEY, J. (1981)
PRESTON, G. N.

Africa,

'First

Tervuren

PRUSSIN, L. (1968) 'The architecture of Islam in West Africa', AA, 1 (2), 32-55, 70-4
PRUSSIN, L. (1970) 'Sudanese architecture and the Manding', AA, 3 (4), 12-19
PRUSSIN, L. (1980) 'Traditional Asante architecture', AA, 13 (2), 57-65, 78
PRUSSIN, L. (1981) 'Building technologies in the West African savannah' inLeSol, la
Parole

et

I'Ecnt: Melanges

Mauny,

Paris, vol. 1, pp.

227-45

RAGGHIANTI-COLLOBI, L. (1968)Bnm;i Museum: London, New York

RANDLES, W. G. L. (1968) L'Ancien Royaume du Congo, Paris

ROSENWALD,

figures', AA, 7 (3), 26-31, 92

J. B. (1974) 'Early Kuba
(1977) 'The iconography of Asante sword ornaments', AA, 11(1), 16-25, 90
ROSSBACH, E. (1973) Baskets as Textile Art, New York
RUDNER, J. and RUDNER, L. (1970) The Hunter and His Art, Cape Town

ROSS, D.

221

ART HISTORY IN AFRICA

RYDER,

A. F. C. (1964) 'A note on the Afro-Portuguese

ivories',

JA//,

5,

363-6

E. (1916) Time Perspective in Aboriginal American Culture: A Study in Method,

Ottawa (Reprinted in D. Mandelbaum, Selected Writings of E. Sapir, Berkeley,

SAPIR,

California, 1963)

SCHACHT,

J.

musulmane

(1954)

'Sur

I'Universite d'Alger, 11,

SCHACHT, J.

(1957) 'An

Orientalis, 2,

SCHMIDT,

les

a travers le Sahara',

diffusions

des formes d'architecture religieuse

I'Institut de Recherches Sahariennes de

Travaux de

11-27

unknown type of Minbar and its historical significance', Ars

149-73
The Ongins of Iron Smelting

P. (1981)

in Africa: a

complex technology

Tanzania, Providence, Rhode Island (Brown University)

SCHMITZ, C. A. (ed.) (1967) Histonsche Volkerkunde, Frankfurt
SEBAG, P. (1965) The Great Mosque of Kairouan, New York

in

am Main

SHALOFF, S. (1970) Reform in Leopold's Congo, Richmond, Vermont

SHAW, T. (1970a) 'The analysis of West African bronzes: a summary of the evidence',
Ihadan, 28, 80-9
T. (1970b) Ihgo Ukwu, 2 vols, Evanston, Illinois
T. (1978) Nigeria: its archaeology and early history, London
SHINNIE, M. (1965) Ancient African Kingdoms, London
SIEBER, R. (1980) African Furniture and Household Objects, New York
SILVER, H. S. (1981) 'Calculating risks: the socio-economic foundations of aesthetic
innovation in an Ashanti carving community'. Ethnology, 20 (2), 101-14
SIROTO, L. (1977) 'Njom: the magical bridge of the Beti and Bulu of southern

SHAW,
SHAW,

Cameroons', AA, 10 (2), 38-51, 90

L. (1980) Review, AA, 14 (1), 76-9
L. (1981) Review, AA, 14 (4), 80-6
SOURDEL-THOMINE, J. and SPULER, B. (1973) Die Kunst des Islam, Berlin
SPEAR, T. (1981) Kenya's Past, London
SPRADLEY, J. P. and McCURDY, D. W. (1975) Anthropology: the cultural

SIROTO,
SIROTO,

New York

perspective.

The Stone Images of Esie, Ibadan

D. (1980) 'Some thermoluminescence datings of terracottas from the
inland delta of the Niger Mali' in B. De Grunne, Ancient Terracottas from West
Africa, Louvain-la-neuve, pp. 276-82

STEVENS,

P. (1978)

STONEHAM,

TAGLIAFERRI,

A. and

HAM MMACHER,
and Guinea,

carvings from Sierra Leone

TALBOT

RICE, D. (1965)

TERRASSE,
TERRASSE,

London

H. (1932) L'art Hispano-Mauresque des Origines au

H. (1968) La Mosquee al Qaraouiyin a Fes, Paris

THOMPSON,
Biebuyck
120-82

Islamic Art,

A. (1974) Fabulous Ancestors: stone

New York
Xlir Siecle,

Paris

R. F. (1969) 'Abatan: a master potter of the Egbado Yoruba' in D.

and Creativity in Tribal Art, Berkeley, Cahfornia, pp.

(ed.). Tradition

R. F. (1973) 'Yoruba artistic criticism' in W. D'Azevedo (ed.). The

Traditional Artist in African Societies, Bloomington, Indiana, pp. 19-61
R. F. (1974) African Art in Motion, Los Angeles, California

THOMPSON,

THOMPSON,
THOMPSON,

R. F. and

CORNET,

J.

(1981) The Four Moments of the Sun,

Washington

TUNIS,

L. (1981) 'The Benin Chronologies',

AA,

14 (2), 86-7

(1948) 'Classification stylistique du masque Dan et Guere de la

Occidentale (AOF)' in Mededelingen van het Rijksmuseum voor
Volkenkunde, Leiden, vol. 4

VANDENHOUTE, J.
Cote d'lvoire

222

REFERENCE MATERIAL

VAN DER MERWE, N. (1969) The Carhon-14 Dating of Iron, Chicago, Illinois
VAN DER MERWE, N. (1980) The advent of iron in Africa' in T. A. Wertime and
D. J. Muhly
463-506

VAN NOTEN,
AT,

(eds).

The Coming of The Age of Iron,

New

Haven, Connecticutt, pp.

F. (1972) 'La plus ancienne sculpture sur bois de I'Afrique centrale',

18 (3/4), 133-6

(1955) 'Initiation rituals of the Bushong', Africa, 25, 138-53

(1968) 'The use of ethnographic data as sources in history' in T. O.
Ranger (ed.), Emerging Themes in African History, Dar es Salaam, pp. 97-124
VANSINA, J. (1969a) 'The bells of kings', JA//, 10, 187-97
(1969b) 'Du royaume kuba au "territoire des Bakuba'", Etudes
VANSINA,

VANSINA,
VANSINA,

J.

J.

J.

Congolaises, 12 (2), 3-54

stad in Centraal Africa: Nsheng', AT,

J. (1976) 'Een negentiende-eeuwse
47-56 (A 19th-century town in Central Africa)
VANSINA, J. (1978) Children ofWoot, Madison, Wisconsin
VANSINA, J. (1979/1980) 'Bantu in the crystal ball'. History in Africa, 6, 287-333; 7,

VANSINA,
22,

293-325

VINNICOMBE,

VOLAVKA,

P. (1976) People of the Eland, Pietermaritzburg

Z. (1979) 'The study of African art as an art historical discipline' in 25th

International Congress of History of Art (published proceedings),

Bologna

VOLAVKA,

Z. (1981a) 'Art and metal: review and preview of research strategy in

African art history' in D. L. Ray, P. Shinnie and D. Williams (eds). Into the 80's:
Proceedings of the Xlth annual convention of the Canadian Association of African
Studies,

Vancouver,

VOLAVKA,
VOGEL,

S. (ed.)

WANNYN,

vol. 1, pp.

149-63

Z. (1981b) 'Insignia of the Divine Authority',

(1981) For Spirits and Kings,

A A,

14 (3), 43-51, 90-2

New York

R. (1961) L'Art Ancien du Metal au Bas-Congo, Champles

WARREN, D. M. and KWEKU ANDREWS, J. (1977) 'An ethnographic approach to

Akan

art

and

aesthetics'.

Working Papers

in

the

Traditional Arts (Aesthetics),

Philadelphia, no. 2/3, p. 19

O. (1970) 'Metallurgische Untersuchungen der Benin-Bronzen des
18, 71-153
Museum fiir Volkerkunde Berlin', Baessler Archiv,

WERNER,

NF

K. (1965) Coptic Art in Early Christian Egypt, New York [Trans, from
German by J. Carroll and S. Hatton)
WILLCOX, A. R. (1963) The Rock Art of South Africa, Johannesburg

WESSEL,

WILLETT,
WILLETT,
WILLETT,
WILLETT,

F. (1967) Ife In the History of West African Art,

F. (1971) A/ncan Art: an introduction, London

London

F. (1978) 'An African sculptor at work', AA, 11 (2), 28-33, 96

and PICTON, J. (1967) 'On the identification of individual carvers: a

F.

study of ancestor shrine carvings from

Owo,

Nigeria',

Man,

2 (1), 62-70

WILLIAMS, D. (1964)'Theiconologyofthe Yorubaeffan0^fcom',A/rica,24, 139-66

WILLIAMS, D. (1974) Icon and Image, London
WILSON, T. H. (1979) 'Swahili funerary architecture of the north Kenya coast' in J.
De Vere Allen and T. H. Wilson, Swahili Houses and Tombs of the Coast of Kenya,
London, pp. 3346
WISSLER, C. (1926) The Relation of Nature to Man in Aboriginal America, New York
WOODHOUSE, H. C. (1977) 'Bushman rock paintings', AA, 10 (4), 21-2, 88
WOODHOUSE, H. C. (1978) Rock Art, Cape Town

223

,**'

INDEX
Figures in roman type
indicate references in

106-7,722, 123, 124,

the text; those in itahcs

Algiers, 9

references in the

Almohad dynasty, 124,

743,210,211-13

captions.

art,

major, minor, 123-4

narrative, 110

124

20

art trouve,

141,

Almoravid dynasty, 210

artists,3,4,26, 56,77, 116,

132, 159
aesthetics of, 87, 102-3,

Abassid Caliph, 154

Alvare de Verre, 34

Abba Lebanos, church, 76

Amen

ablak, 59

aminoacid chronology, 38-9

identity, 27, 138

Abuja, 180, 184

anachronism, 203

and

account books, 34

AnatoUa, 99

acquisitions, 27, 29, 33, 34

ancestor figure, 31, 47, 50,

Adam,

121, 130, 132, 133,

Khopshef, 69

additive sculpture, 57, 61,

130, 157, 176, 210

Andalusia, 92-3, 83, 160

92, 144
aesthetic, 87, 123, 129,

130-3, 136, 147

African Art, journal, 29,192

Afro American an, 11

Afro Ponuguese

art, 14, 21,

Ashir, 9

142, 210

Angola,

3, 14, 75, 78, 34, 65,

29,46, 103, 104,

northern, 36, 42, 67, 72

187,205

Akan
and Middle Niger, 33,
119, 159, 166,767,
171-2, 203

goldweights, 75, 112,

162-3, 201-3, 202

Akjoujt mines, 7, 8, 75

Akota people, 32-3, 32

al Aqmar mosque, 98-9
al

Bakri, 34, 54

al

Hakim mosque, 99

al

Qahira, 154

Alexandria, 8, 160
Algeria, 9,

10,93,776, 129

southern, 33, 61, 93,

89-92, 95, 143-5,

187, 191

Attabubu, 72, 172

attribution, 89-92
d'Aulvekerque, Jehan, 34

73

Aghlabid dynasty, 25, 154,

117, 175,

201
Asante, 11,73, 175, 176

atelier, 70,

central,

194

style, 78, 89, 111

travelhng, 29, 36, 160-1,

anecdote, 130

34,35,40, 185, 187-90

Agadez, 8, 12, 60

Aka pygmies, 75, 130

Akan people, 13,46,72,83,

and

Andreas, Saint, 100

angel,40,45, 102, 110,775,

Afghanistan, 119

society, 44-6, 50-2,

102-3, 121, 136-40

84-6, 86,87, SS, 106,

113

136-40

Axum, 8,76,77
d'Azevedo, Warren, 139

161, 193, 205

Anorewos, painter, 775

Bafut, 29

antelope, 7'^6, 151

Baha, 786

Anthony, Saint, 158

Baibars, mosque, 98, 99, 154

applique cloth, 72

ballet,

apprentice, 50, 143

Bakota, people, 31-3, 32

Arab, 10, 112, 133, 156, 160

Bamako,

arabesque, 116, 124, 129,

133, 145, 163, 164,213

Arabic script, 116,776, 117,

Bamana, people, 73, 29, 146,

Bamba

province, 32, 32

Bambara

145
arch, 76, 102-3, 145-6, 164,

see

Bamana

Bamessing, 29

Bamileke people, 75, 123

199

124,725,127,

148, 154, 755

3,

kingdom, 77

cliffs, 12,36,
38,38,67, 187,

189-190, 789, 193

Bangwa

126, 185
art, definition of,

Bamum

Bandiagara,

72,21,24, 29, 78,

101,102,106,109,121,

Ardra,

72, 146, 187, 789,

191

175

Arabia, 19

islamic, 57,

128

1-2

people, 73, 29, 51,

132, 139

225

INDEX

basilica, 76, 77,

see Ethiopia,

Byzantium, 95, 116, 135,

barkcloth, 70, 86

clientele,

164, 194

82

coiffure see hairstyle

basketry, 42, 70, 72, 73

collective representation,

Baule, people, 13, 106

Cabinda, 75, 112

Baumann, Hermann, 170-1,

Cairo,*, 21, 22,98-9,

121, 140

colour, 86, 123, 128, 148,

154-5, 172, 176

177
beads, 47, 72, 104, 186

calabash, 56, 83, 101, 108

beauty contest, 130-2

calUgraphy, 1234

Befale, 151

Cameroon,

10, 75, 14, 75,

3, 12, 13,

185, 187, 203

Cape, province,

plaques, 95-6,96, 97,

Congo, 123, 72*, 130, 160

59,59, 71, 113, 154,

97, 147, 178-85, 186

Benin, Republic, 2, 5, 191

14, 162, 180, 184

133, 203

convex

Carthage, 7

Coptic people, 7, 10, 46, 51,

cast,

catalogue, 3, 21, 24, 34, 40,

lamps, 119, 159, 171-2

painting, 70, 110, 111-12

185

cephalomoiph cup, 110,

sculpture, 39, 117, 77S,

147

Bigo, 16, 17

148, 750, 164-5, 765,

Binji people, 765

176

West

body painting,
Boer people,

3, 17, 73,

Natal, 17, 191, 792

Chad, lake,S, 11,72,75,72,

75, 76, 162, 210

127

17, 18

charm, 41, 47, 50, 68, 83,97,

Brandberg, mountains, 108

Bravmann, Rene, 129, 163

Brazzaville, 31, 760
Breuil,

Abbe, 108

brick, 24, 25, 57, 124, 724

Brong, 172
building, prefabricated, 59
Buli, 75, 36, 89, 90, 176

Bullom people, 13, 35

see also

Afro Portuguese

775, 752, 160-1, 202,

207, 208, 210

China,22,23, 159,795, 199,

people, 75, 29, 103,

104, 106, 110-11,205,

9-^,

109, 157,

4S, 97, 160

136-73

Crete, 108

aesthetic, 130-2

or relatedness, 142, 144,

styhstic, 176

crocodile, 117,775, 129

158, 160

Bunmdi, 76, 172

bwami association, 46
Bwari, 180, 184

church architecture, 47,

226

shell, 47,

creativity, 45,

164-5, 168, 170-2, 195

208

chronology, 33-40, 74, 95,

Bwiti cult, 123

cowry

36, 46, 135, 147,

204

criteria

200, 203

Christ, Jesus,

counterfeiting, 142-3

coun an,
193,

Chibinda Ilunga, 26, 103,

104, 106, 110, 208

Chokwe

identity, 31, 41, 111,

cotton, 62, 70,77, 104, 166

Charles the Bold, 114

33

84-5,101,117-19,126,
127, 168, 201, 204
157, 196

Bolony, 65
9,

Cordoba, 9, 148
costume, 26, 53, 62, 65, 68,

and

75, 163, 184

chandelier, 114

Bolombo, 29, 54
Boudenib,

77

163

Africa, 11, 177-87,

181,183, 188

156

textile, 52,

copy, 141-3, 154, 156, 159,

ceramic sculpture

blacksmith, 41, 51, 52, 61,

95, 138, 146-8, 156-7,

architecture, 59, 82

Betsi people, 75, 85

9-^,

172-3, 177

catacomb, 76

Catholic, 1.6, 194-5

bishop, 36, 44, 45, 94-5,

style,

160, 213

open, 61

Berber j)eople, 44, 73, 124

birth, 41

convention, 45, 78-83, 129,

carbon 14 dating, 56, 38, 39

carving technique, 65-8, 65

110, 117,201, 210

royalheads, 39, 95-9, 96,

747,201
concave style, 85, 172-3, 177

176, 187, 189

ivory, 19, 34, 50, 67, 68,

117-18, 204

6, 10, 18, 18

Cape Lopez, 160

Cape Town, 6, 7S, 61, 62
capital,

court, 36, 46, 47, 50

175, 194

composition, 103-6, 120,

canvas, 83

19, 26, 75, 105, 180,

commission, 44-7, 102, 123,

138, 140, 155-6, 160,

82, 123, 132, 175

canon,84, 127,133, 175,180

Belgium, 91, 172

bell, 114, 161-2

style, 81,95, 100,135

and technique, 68, 69, 70,

73, 79, 146

grasslands, 72, 29, 31,51,

Belem, 26

Benue,

162, 190,206, 212

and

29, 45, 139

Begho, 12, 166, 189

Begram, 119

Benin, kingdom,

Nubia

44-7

Cross River, 10, 11, 72,68,

81, 99, 100, 103, 129,

99, 164, 178, 194

1

14,

126, 140, 163, 165, 204

134

crown, 39, 42, 42, 161

INDEX
culture, defined, 121, 133-5

ancient wallpainting, 6,

culture area, 176

68,69, 78-81, 82, 108,

cuneiform, 26

109, 168, 769

curiosity cabinet, 19, 34, 161

currency, 160, 193

cut pile cloth, 72, 77, 766

islamu, 8, 10, 64, 99,

94,95, 109, 110, 130,

135, 147, 156-7

Fatimid dynasty, 21, 51,

97-9, 138, 154-6, 164,
194

154-7, 155, 203

see also Coptic

people

Fes,9,46,77, 114, 176,210,

211,212

Einstein, Carl, 19

Dahomey,

3,

11,75, 117,

123, 191

filigree,

Ekoi people, 75, 21, 81

fish-legged figures, 117,779,

Daima, 186

ekpe association, 134

Dan

people, 75, 31, 45, 132,

elephant, 129, 210

205

emblem, 44,54,

dance and mask, 47, 53, 68,

83, 126-8, 148, 161

Dark

Philip, 95, 96

dating, 20, 21, 96, 174

see also

chronology

55, 159,

164, 193, 210

on sculpture, 84, 87, 111,

201, 204

120
folk art, 46, 163

Fontem, 29
forgery, 20, 22, 24, 26, 27,

55

FrancevUle, 80, 81

embroidery, 71, 726, 167

Frazer, Douglas, 118

emotion, 108-9, 110, 114,

France, 19, 108

Daura, 72, 117

116,120,121,127,132,

David, king. 111

211

Debra Berhan Selassie, 112

decoration and meaning,

England, 18,

Frobenius, Leo, 19, 29, 57

54
19, 95, 159,

163

Fulani people, 75, 108, 163

function and use, 41, 44,

52-3

epiphenomenon, 157-8, 196

114-17

75-6, 166

Eket, 99-100, 103

Dekese, 149

Eshu, God, 30, 101

funeral, 41, 199

dendrochronology, 36,

Esie, 72, 74,

17^81, 185
Eskimo people, 28

Gabon,

38-40

de Vere Allen, James, 200

devU,

3, 102,

112

distributions, 164-73, 193-4

documents, written, 21-9,

117, 203, 210

76, 76-7, 82, 92, 102,

103, 122, 138, 146, 164,

135, 138, 201

194

people, 75, 67, 84,

109-10, 130, 177, 210

doll, 46, 50,

61,62,65

dome, 57, 76, 82

Egypt, 755, 156

81,87
northwest, 26, 29, 84-5,

86,87, 133
south, 45, 123, 135, 204

Gao,S, 72,789, 189

painting, 36, 50, 83, 92,

Garamantes, people, 77

112, 114-16,775, 122,

Garrard, Timothy, 63, 201

129, 138, 141, 160,

gate, 44, 46

194-5, 204

George, Saint, 117,778,127,

142

sculpture, 17, 67, 122

ethnic group, 27, 29-33,

drawing, 34, 211

ethnicity, 205

drum, 44, 47, 4,'/9, 54,59,

ethnocentricity, 123-4

45-6,73,85,164,175-7

Gere

145-8, 156-7, 174,

187, 193, 195

ethnonym, 21, 27, 29-33,

50, 85, 173, 175-6

Wee

see

Ghana, empire,

11, 75, 34,

117, 193

Ghana, Republic, 75, 129,

163, 191

ethnographic present, 4

111
drift,

northeast, 31-3, 50, 80,

miniatures, 36, 124

Maghrib, 77, 92, 95,

114, 148, 194,210,211
dragon, 102, 117, 142,203
in

14, 75, 177

coastal, 36, 160, 161

architectiu-e, 36, 42, 57,

33-6,40,53-5,72,132,

Dogon

in

Ethiopia, 7,5,10,76,33,51,

Gibraltar, 114

Gio

see

Dan

Duala, 72, 757

evU, 129-33

glass, 64, 68, 126

Dupre, Marie-Claude, 127

Eyo, Ekpo, 178

Goa, 204
goal, 52-3, 57,

dyeing, 68, 70-3, 72, 77

95,99,724, 156
Fagg,WUliam, 31, 54-5,95,

fagade,

eanh goddess,

146, 181

edict of Milan, 7

Egypt
ancient, 6, 7, 26, 38, 45,

63,67, 82, 111, 117,

122, 135, 147, 172, 203

96

Fang people, 75, 26, 29,

84-5, 86, 95, 133
Faras, 8, 64
wallpainting, 36, 45, 52,

God,53, 117,

99

118, 130, 154,

208, 210

Gok, madrasa, 99
Gola people, 75, 138-9
gold art objects, 34, 61, 74,
163

goldtrade, 11,17, 160, 162,

227

ART HISTORY IN AFRICA

166, 193, 201, 203

goldweights, 52, 112-13,

117, 144, 147-8, 162-3.

icon, defined, 101-103, 757

Istanbul, 10

identification, 21, 22-4, 28,

Italy, 23, 34,

166,201,202, 203

29, 106, 110-11, 136

Goliath, 112

idol, 34, 36, 129, 163

Gombrich, E. H., 78, 112,

Ifa,

136

ifa

god, 3

106, 132, 161, 189-90,

78, 101, 102, 109, 121,

Goundam,

126, 185

12, 187, 189

Ife,

Guere

see

129, 177-85, 756, 204

Igbo people, 75, 31

Igbo

Guernica, 140
guilloche, 117,

167-8

Guinea,75,44,67, 168, 177,

190

Guro people,

75, 79, 81

ivory as mediiun, 56, 65,

67-8, 74

Igala people, 75, 110, 161

Wee

201, 205

72, 64, 74, 81, 95-6, 111,

Griaule, Marcel, 110

Grottanelli, V., 34

Ukwu,

Japan, 57, 204

Jebba, 72, 178, 180, 185
jebel

72, 61, 63, 75,

75, 178-80, 184, 186,

Jenne, 11,72,187, 189,759,

203

193

Ijebu Ode, 180, 185

Jerusalem, 8, 157, 160

Ilesha, 180, 185

jewellery, 61, 84-5, 110,

164, 167

hairstyle, 3, 31, 73, 84-7, 97,

127,204

Uwainat, S, 107

Jemaa, 181, 184

illumination, 36, 124, 144-5,

110, 111,

126, 144, 189, 204

joinery, 67

imam, 44, 154

Josephine, nickname, 106

Hallen, B., 132

imbol design, 117, 167

Juda, 129

Hasan, Sultan's mosque, 99,

Independent Congo

Jukun people,

India, 7, 19, 76, 119, 167,

204

people, 13, 117, 162

knot, 30, 166-8, 767

161, 193, 196, 199, 200

Indonesia, 19, 167

145, 146

people, 75, 50, 88,

89, 95, 170, 176

industrial revolution,

65, 108, 127

117, 145

Hispano-maghrebi

art, 10

historical critique,

53-5

hodegithria,

Chokwe, 106, 110

Kuba, 47, 113-14,775,
205, 206
innovation, 22, 45, 46, 52,

madonna, 775

Holo people, 75, 208, 209

holy men, 93, 122, 142,
211-13

inscription, 21, 25, 33, 36,

154,157,175,193,204,
210
interlacustrine area, 7, 17,

horro vacui, 147

Horus, god, 117, 118

Hum

101
interpretation, 106-10, 136,

203, 205, 208

people, 212

hunter, 18, 26, 103, 104,

105, 208
ibadi Muslims, 163
Iberia,
Ibi, 72,

Iran, 117, 164, 165

184

river, 180,

Kafanchan, 184

Kambundi, village, 66
Kanem, state, 11, 75
Kankunde village, 89
Kano, 72, 167
Karkur Talh, 107
Kasai, region, 14,75,25,81,
89, 161, 164, 172, 191
east, 83, 90, 91, 133, 139,

204
Kateba,

village, 89,

90

Katsina Ala, 72, 184

Ke,

11

Keenge,

village,

772

Kenya, 76, 17, 53, 82, 196

Khoi people, 62

Khosa people, 62

Iraq, 156, 194

Ireland, 117

Kinshasa, 207, 212

iron sculpture, 61, 68, 74,

83, 139

Iroquois people, 167

Ibibio people, 129

Isis,

Ibn Batutta, 34, 54, 190

Islam and figurative art, 126,

228

Cabinda
9, 134, 162

kidumu, mask, 52, 127, 148

Kina, district, 22-3, 23

34
162

see also

Kabyha,
Kaduna,

109, 111, 138, 201

institution, 46-50, 134-5,

126, 146, 151, 201

210

Kabila, figure, 90, 106

133-4, 136-58, 195

horn, 79, 85, 103, 110 111,

ivory, 34, 166,

initiation, 41, 45, 46, 47, 51,

Herodotus, 7
hieroglyph,69,82, 109, 114,

75, 162

Kabinda, 42

Indian Ocean, 6, 14, 17, 160,

Hellenistic art, 46, 119, 779,

Hemba

State,

55

725, 143, 155, 156

Hausa

133

Ivindo river, 32, 32, 33

Ivory Coast, 3, 31, 79, 81,

divination, 2, 3, 29, 30,

Gondar, 112
Greek, 1,7, 56, 81,135, 157

129, 163

iconatrophy, 203

goddess, 22, 69, 172

Kisi people, 75, 187, 190

Konge kingdom,

14, 75, 22,

25,-^2,46,158,766,210
people, 61

Kongolo, 75, 89

INDEX
Konya, 156
Koryak people, 28
Kota people, 32,52, 33, 50
Kuba kingdom, 14, 75, 25,
42,46,47,45, 51,54,

Mali, empire, 11, 75, 34, 54,

Leuzinger, Elza, 177

160, 187-9, 190, 193

Lezine, Alexandre, 154

Liberia, 75, 138, 139, 189,

eastern, 31,45, 47, 51, 53,

Likana, village, 72S

765
decorative art, 116, 117,

123,726,138,147,156,
167-8
initiation,

Limpopo

205
royal statues, 36, 148, 752,

154, 157, 175

other statuary, 74, 75,

160

Margais, Georges, 92

Loange,

Marianos, bishop, 94

river,

coast,

75,29,39,42,130,757,
159, 160, 210

wax

river,

person,

region, 14, 15,66,

67, 81, 89, 191, 208-9

casting, 11, 61-3,

marriage, 31, 41, 73, 126,

mass, 57, 61, 122

mass production, 102

202

master, individual, 87-8, 89,

92,94, 138, 144, 176

Luanda, 75, 29, 31

Luba, kingdoms, 14,75, 193

master of the madonna, 94

Luba-Hemba, people,

Matamba, kingdom,

75,

mat, 73, 138

styles, 29,

57

89-92,

14, 75, 104,

Matara, 8, 77
Mauritania, 7, 75

Mayi Ndombe,

lake, 75,

128, 161

161
spirit, 42
Lyndenburg, 7S, 792

Mayombe,

lamp,64, 119, 159, 162, 163,

171-2, 198-9

Maack, person, 108

Lamu,

Mabiinc, person, 138

Mbeti people, 32, 32

Mbole, people, 151

Mack, John, 72

Mecca,

Madagascar, 10, 76, 17, 18,

Mekery, people, 85

LaUbela, 5,42,76,702, 164

Lamb

Venice, 72

76, 196-200, 211

language and area, 32-3, 85,

170, 175, 195

Lavachery, Henry, 177

lead, 61, 63,
leather,

Lusunzi,

19, 92, 161,

203

Madina, 148

96

33,62, 68, 72,88,

madonna,94,114,

Lega people, 75, 46, 129

Legba, god, 3, 30, 101
Lele people, 75, 47, 49, 765

157, 195,

Melchite church, 157

Mehka,

9, 122-3, 722

memento, 143, 151

memory, 101, 121, 143

93, 124, 148

Malebo pool, 159

Leopoldvilie, 211

9, 46

Maghrib,S, 9, 10,25,44,72,

Leroy Jules, 82

66

leopard, 123, 151

Meknes,

199

Melusine, 119

Leptis Magna, 8, 160

Lembwa Basuku,

7, 8, 24, 160,

madrasa, 99, 147, 155-6

Mahdi, 154
Mahdiya, 9, 97-9, 98, 154
Majabat al kubra, 75
Majunga, 76, 19
Malawi, 76, 17, 18
Maldive Islands, 160

Lem, F. H., 67
Lemba, 211,272

region, 166

Mbailundu, 75, 29
Mbeengi, village, 772

Melanesia, 19

208

145, 168

75,

208-10

176

Lunda, empire,

Lagos, 72, 63

Marrakesh, 9, 143, 144

Masai, people, 76, 117, 142

148

90, 133, 161, 176

lacuna, 187, 191

Marinid rulers, 211-13

75,75,77,172,189-91,

Luba,
laboratory, 36-9

17, 18

197, 199

Loir, Helene, 72

Lualaba, lakes, 172

172

Kwango,

164

Loango kingdom and

lost

Kufic script, 776, 145

Mangbetu, 58, 59
Maniema, region, 14, 75,57
74
Mapey, village, 113

Mapungubwe,

165

Kusu people, 89
Kwabena Bonda,

rulers, 10, 59, 82,

hteracy, 33, 117, 134, 138

Lokenye,

726, 166

Mamluk

mappula, 114-16, 775

183, 199

143,151,755,157,764,
textiles, 70, 77,

Mali, empire and

Lisbon, 22

111,772, 160-1,205,
206, 207, 208
other carving, 65, 140,

river, 17, 18, 135,

linguistics, historical, 177,

113-14, 773,

see also

99, 117, 154-6

160, 203

cups, 110, 148, /5(?, 164,

17, 146,

Dogon

85-7, 133, 205

Libya,*, 10,44,82,92,707,

164, 765, 176

146, 188

190

56, 61, 126, 135, 139,

Mali, republic, 73,

Menas,

Saint, 160

Menna, person, 109

Menzel,

Brigitte,

72

Merina, people, 76, 19

Messiah, 154

metalwork. Islamic, 33, 52,

126, 144, 203

229

ART HISTORY IN AFRICA

Meuse,

Namibia,

river, 176

3, 6, 17,

78, 108

Ntumu

people, 75, 85

Michael, archangel, 45, 142

Nasarawa, 184

Nuba,

migration, 18-9, 161, 193

Natal, 7, 17, 18, 191

Nubia, 7,8,

mihrcb, 24, 46, 124, 140-1,

nature

148, 199, 210, 211

spirit,

architecture, 36, 45, 52,

Mijikenda people, 76, 53

Ndengese, 75, 148-51, 149

minaret, 25, 45, 59-61, 98,

negative evidence, 22, 55,

124-5, 143, 144, 148

70, 94, 95, 122, 130,

Ndumu

missionary, 14, 22, 53, 55,

Neptune, 118-19

Molet, Louis, 203

people, 15, 31, 75/

monophysite, 157

monument,

33, 36,

14, 122,

object, atypical, 22, 26

Ngeende people, 772

Ngongo people, 65, 772
Ngongo ya Chintu, person,
89,90

Ngumba,

138, 201

135, 141, 194,208,210

Nupe, 73, 129, 163
Nzinga, Queen, 208-10,209

176

Ngabe, 160

195

Mongo

people, 32, 80

Neyt, Francois, 87, 95, 100,

158

59, 82, 109, 135

wallpainting, 36, 45, 52,

203

minbar, 44, 163, 208

model, theoretical, 177-93,

10, 51, 64, 74,

157, 168, 203

101-2, 113,

157, 207, 208

S, 117

foreign, 159-64, 171-2,

203-4
objet trouve, 27, 41,

Ogboni

75, 85

181, 185

OgotemeUi, person, 109-10

Mopti, 12, 187, 189, 189

Niani, 72, 189, 759

Morocco, 9, 33, 46, 93, 134,

niche, 198-9

Ogowe,

Nicklin, Keith, 100

oikoumene,5,6,

147, 767

mosques, 46, 725, 140, 141,

143, 148,211-13

Niemba,

morphological analysis,

Niger river bend, 36, 61,

83-9, 99-100, 165

mosques, 21, 24, 33, 41, 53,

82, 102, 204
minaret, mihrab

East African, 199

North African, 45, 116,

125, 126, 134, 140-1,

143
Fes,

Qairawan, portal
Africa, 60, 61, 123,

163

Mossi people, 13, 190

motif, defined, 102-6

Mozambique,

6, 76, 17, 18,

187
river, 103,

105, 161, 178, 180

59,60, 163, 177, 191

mud sculptiu-e ,17,61, 64-5

122-3, 127, 757

Muhammad, prophet, 7,112

177, 187-91, 188, 189

oral tradition, 48, 95-6, 139,

204

Niger-Imo area, 31, 61,

Orange Free

State, 17, 18, 135

65,65,73,75,142,172,

Orvieto, 34

184

Ottoman empire,
ownership, 47-50

181, 191

Owo,

southwestern, 3, 10, 11,

25,29,30,74, 82,97,

10, 82, 92,

93

117,129,162,163,167,

72, 779, 178-85, 186,

204

Oyo, 72, 73, 178, 180

117-20, 119,779, 130,

Padua, 22

Nile, river, 6, 72,93, 117

painting on glass, 64, 126

Nile valley, 213

painting and interpretation,

102, 106, 108-9, 709,

Nioro, 75

Nkuvu, chiefdom, 89
Nok,7,72,64, 177-84,

110, 112, 157,201

186,

painting, rephcation, 142,

ethnonym

painting, style, 78-81, 87,

144-5, 148, 204

191

nommo,

music,2, 127, 128, 139, 164

Northern, Tamara, 36

Mvae, people, 75, 85

Nsoko, 166

230

poetry

Nigeria, 13, 15, 168, 177-87

nomenclature

Ngumba

an, 126, 127, 129, 140

see also

oral data, 33-6, 55, 201

111

see

89, 161, 176

46, 162, 163, 166, 172,

Munza, king, 58
muqamas, 77, 755, 210

Mvimibo

people, 75, 85

Old Oyo, 180, 185

Olokun, god, 118
oral

142, 154, 167, 178-85

19, 163

mud architecture, 56, 59-61

mummy.

Okak

Olbrechts, Frans, 33, 39, 84,

160, 191

Lower

7, 33, 36, 46,

134, 159, 160, 201

northern, 7, 11, 72, 75,

Sahara, 145

West

Niger, country, 13, 60

Niger,

river, 75, 29, 161

70, 110, 121, 122, 123,

Niger, Upper middle river,

minbar

see also Cairo,

88

Niger Delta, 67, 114, 161,

mosaic, 70, 167

see also

area,

130

association, 74, 154,

see

spirits,

36, 38

Nsongo people, 151

92,94,115, 129, 147

painting, styhstic series, 97,

147
painting techniques, 25, 64,

INDEX
25,77,92,114,148,194

68-70, 69, 126

palace, architecture, 58, 59,

59, 82, 122, 123, 163

Qait bai, mosque, 21

Banu Hammad,

Qal'a of the

palace, identification, 42-4,

queen, 208-10, 209

47

123, 164
Sahel, sculpture, 6, 11, 51,

99

9, 77,

145
Sahel, architecture, 59, 77,

67, 117, 130, 174, 175,

191, 193

Palestine, 95, 135

quilting, 72

Pate, 196, 199

Quran, 41

Sakalava people, 76, 18

Quttubiya mosque, 143

Saladin see Salah al-Din

analysis of, 25, 26, 27

Rabat, 9, 725, 144

salt cellars, 34,

patrician, 46-7, 150, 156,

raffia material, 56,

patina, 3S, 66, 67,

74,79,

Salah al-Din, 10, 156

132, 133

157, 196, 200

71,

patron, 44-7, 52, 102, 136,

S, 766

Sangha, river, 75, 123

raffia cloth, 70, 83,

Ramses

138, 145, 156

113,725,205,272

55
San people, 62, 70
Sande association, 111

58, 68,

III,

Sankuru

69

patronage, 41, 44-7

Raphael, archangel, 45

Pecile, Attilio, SO, 81

Red

Pende people, 75, 81, 89,

reUef, 118, 122,7-^7

Santa Maria Maggiore,

200

church, 775, 116

Sape people, 75, 161

reliquary figure, 31-3, 50,

164

performing

sea, 160,

arts, 121,

126-9,

80, 81-2,

S7,S6

Sassanid empire,

Renaissance, 6, 46, 81, 119,

208

repertoire, 45, 127, 129

7,95
perspective, 78, 81, 82-3

scarification, 31, 56, 73, 84,

replica, 102, 136, 757, 140,

142-8, 161-4, 203

Phoenicia, 7
Picasso, Pablo, 140

17

Satan, 112

Scandinavia, 117

142

Perrois, Louis, 84, 85, S6,

river, 75, 29, 54,

148, 153, 204

prime work

see also

85,97, 111, 117, 126,

148, 204
Schacht, Joseph, 163

Picton, John, 72

restoration, 26, 55, 753, 192

Scheldt, river, 176

pilgrimage, 160, 171

Rhine, river, 176

scriptoria,

pipe, 42, 51, 164, 765, 176

Richard

Pitt-Rivers,

Augusus, 19

plaster decor, 33, 116, 198,

199
poetry, 120, 127-8, 133,

Polakoff, Claire, 72

pole sculpture, 17, 52, 67,

portal,

rock

art, 17, 38, 122,

rock

art,

Senufo people, 75, 25, 175,

139

southern Africa,

SevUle, 9, 144

193-*

Saharan, 4, 6, 27, 70, 106,

707, 108, 168, 169, 193
city, 22,

98,

154-6, 755
portrait, 85, 111, 772, 140,

182,20^,209

Shaba, region, 14, 75, 36,

8, 89, 90, 90, 106,133,
160, 161, 170, 172

25, 55, SO,

Roman art, 27, 99,

Shaka, king, 17

111,154,

163, 167, 194

Roman

177, 189

3,

775, 116, 122

Fatimid and

Mamluk, 97-9,

Seljuq rulers, 99, 156

rococo motif, 34, 83

Rome,

177

empire, 7, 77, 82,

77S, 119, 160

Shango, god, 130

Shari, river, 14, 75
shell, 25, 68, 72,
see also

Rovuma,

potter, 51, 64, 138

rug, 33, 124, 124, 134, 199

prime work, defined, 84,

Russia, 117, 133

Shi'i

737, 141, 142, 148-54,

Rwanda,

shrine,

156, 193

Rweej, person, 104

river, 76, 17

76, 172

89, 148

prototype see prime work

proverb, 112-13, 203

Sa'ad, person, 21

muslims, 154-6

42,42,44, 47, 97,

123,757, 160, 161,755

king, 752

Saba, 170

Sidi Aissa, 122

Sadratha, 9, 33

Sidi bel

saint, 110, 111, 157, 158,

Sierra

160, 208, 210, 211-13

Sahara, architecture, 10,60,

Quaitawan, mosque, 9, 10,

43

Shyaam aMbul aNgoong,

determined, 85, 87,

proportions, canon of, 85,

160-1

cowry

Sherbro, peninsula, 12, 35,

Portugal, 17, 34, 158, 203

99-100, 140

145

Sebk, god, 117,778

Sefar, 8, 106

rinceaux, 145

'34,

199-200

of England, 159

II

61,77,92,93,722,123,

Hasan, 776

Leone, 72,75, 34,55,

43,44,67, 111, 161,

187, 189-90, 193
Signorelli,

Luca, 34

231

ART HISTORY IN AFRICA

Sijilmasa, 9

Tanzania, 76, 17,28,29,37

Tripohtania, 93

Sivas, 99

Taqali, kingdom, 117

Triton shell, 65

skeuomorphism, 56, 86-7,

Tarkwa, 72, 171, 172

Tsayi people, 75, 32, 127,

83, 124, 185

Taruga,

11,51, 129, 131, 139,

157, 160, 199-200

Tastevin, Roger, 42

Tsauni mine, 181

Tellem people, 13,36, 37,

Tsogo people, 75, 123, 204

Tswana people, 78, 135
Tumbica, 75, 191

slave,

smithing, techniques of, 25,

38, 38, 187, 190

Tempesta,

61-4
Smuts, Jan, 108
So people, 184
social stratification,

46-50,

184

7, 72, 180,

p)erson, 82

temple, 14, 47, 82, 123, 204

Tunis, 9, 148

Thebes tombs, 69, 108, 109

Tunisia, 9, 10, 44, 64, 160

Thecla, Saint, 747

architecture, 25, 95, 98,

157, 158, 163, 196-7,

theme, defined, 102-6

200, 204, 205, 213

thermoluminescence, 38, 96

Somaha, 8, 16, 167, 168

Thompson, Roben

Songhay, empire, 11, 13,

Thoth, god, 69

F., 132

throne, 24, 44, 161, 191

193

Songo people, 75, 205

Songye people, IS, 83, 90,
97,92, 133, 161
South Africa, RepubUc, 18,
62
Southern Rhodesia, 108
stone, 39, 73, 74, 137

Angola, 29, 34, 106,

92-100, 142, 143, 195

Benin, 95-7, 96, 178, 204

Morocco, 140, 141

subtractive technique, 32,
57, 61, 65-8, 144

Sudan, 5,75,76,17,72,92,

93,94, 117, 124

Sufism, 212-13

Suku people,

75, 66, 67, 89

Sunni musiims, 156, 163

Swahili people, 76, 17, 126,
133, 196, 199-200

symbol, 41, 101, 110, 113,

728, 129,757, 160,203

Ulm, 21, 72
Upper Volta,

Tio, 75, 31, 70, 160

11, 75, 187,

191

urbanization, 11, 194,

196-200

Maghrib, 61, 92, 93,

122-3,722,142,211-13

Vallisnieri, person, 22, 25,

mamluk

27, 29, 34

99, 755, 156

pharaonic, 69, 108, 709

tool,34,41,44,77, 109, 137,

194, 201

Vandenhoute, Jan, 132

vault, 57, 58, 155
Vili,

people, 210

Torday, Emil, 54

vinuosity, 56, 140, 157

tourist art, 130, 757, 142,

visual concept, 101-3, 114,

121-2, 134, 757, 210

145

Volavka, Zdenka, 42

toy, 163

Volta, river, 72, 75, 187,

189, 790, 191, 193

Tozeue, 9, 124
trade of art in Africa, 29,31
33, 159, 168,

203-4

depicted in an, 110, 130,

757

emporia, 17, 171,184, 196

and diffusion of style, 134,

176, 203-4

Tada, 72, 64, 178, 180, 185

Takedda, 75

trader, 55, 139, 171, 194,

Taki, 80, 81

Transvaal, 17, 18, 135,

232

ibn Nafi, person, 10

use and shape, 2, 42

Tabora, 76, 29, 31

191

160 162,

figures, 65, 172

routes, 75, 170, 171

Tanganyika, lake, 76, 77,

river, 159,

171

Uqba

9, 747, 144

Syria, 156-7

Tamelhat, 9, 124
Tananarive, 76, 19

Udegi, 184

TinMal,

see also doll

Suguni, 146

Udo, 95, 180

Uganda, 76, 17

769

tombs, eastern Africa, 19

148, 190

wara headgear, 146, 746

Tibet, 107

143, 755, 191

styhstic seriation, 85,

194
tyi

throwing knife, 168, 170,

98, 124,725, 135,747,

sculpture, 65, 67-8, 747,

99, 124, 154, 160, 164,

Ubangi

110-11,205

Tlemcen, 9, 777, 129

Togo, 13, 67, 191

architecture, 57, 59, 82,

148, 157

201

191-2
tribe as label, 21, 29-33,

175-6
Tripoh, 5, 9, 73

volume

in aft, 61-8, 84-7,

122-3, 133, 136, 148

description of, 90-2, 100,
106, 132, 785

Von Luschan, Friedrich,

19,

95

Wagadu,

75, 117

Wagania,

village, 151

wallpainting, 27, 74, 83,

126, 140
see also Coptic,

Egypt,

Nubia
weaving, 50, 70-2, 77, 166,

766

INDEX
Wee

people, 75, 31,45, 133

southern, 22,

Yola, 184

Yoruba people, 13, 117-20,

welding, 61

WUlett, Frank, 74, 124, 178

Wilson, Thomas, 92

Windhoek, 18, 108

window, 102, 103, 144, 146

119,

US, 167, 186

sculpture, 3, 25, 29, 30,

106, 123, 130, 142,

178-85

witchcraft, 41, 129, 138,139

womanhood,

94, 112, 153,

205-10, 206, 207, 209

wood,

analysis of, 24-5, 27

wool, 70, 72

Woot, person, 113, 116

workshop operation, 50-2

Zaire,/-/,

15,25,49,66,67,

rays, 25,

210

17, 18, 162,

Zambia, 76, 17, 18, 160

Zande people, 75, 124
zawiya, 212-13
Zaytuna mosque, 194
64, 144

212

Ziba people, 28

89, 124, 193

lower, 29, 36, 64, 68, 77,

river,

middle, 24, 31, 65,

73,757, 161

97, 170,

170

zallij tiles, 52,

82, 163, 166, 205, 210

Zambezi, 76,

113, 148-51,749, 176,

eastern, 46, 50,55, 73, 74,

96>,

191

Zimbabwe, ancient,

17, 123,

130, 135, 161-2, 762,

191, 204

Zimbabwe, Republic,

17,

18, 82, 108, 160, 163

Zulu, 17, 18, 62

Yelwa, 177, 180, 184, 186

233

The

history of African

systematic

works of an

method by which an

is still

historical

in its infancy.

This Jjook provides a

understanding of African art can be

achieved and uses numerous photographs and drawings of objects from the
entire continent to illustrate the questions, findings

and arguments

raised.

Professor Vansina considers the identification of an art object, its relation to

its social matrix, the study of its medium and technique, its style and meaning

and

its

relationship to other arts

and

to culture, as well as the creative process

through which such objects come into being. The book places African art
within an historical framework which explains how artistic changes have
taken place.

Art History in Africa

is

lively

introductory textbook which will appeal not

only to students of art history but to

all

those interested in the arts and culture

of Africa.

Jan Vansina

is

one of the most distinguished historians of Africa, and is

numerous

Professor of African History at the University of Wisconsin. His

books include Kin gdoms of the Savanna Oral Tradition The Children of
Woot and The Tio Kin gdom of the Middle Con go: 1880-1892
,

Cover photograph: Akan goldweight of a Sankofa

by kind permission of the Trustees of the

British

bird, a sign for history in

Akan

art,

Museum,

ISBN Q-Sfla-bM3bfl-b

^ii-

Longman

9 '780582"643680